Deloitte Uk Ido Playbook 2022 Final1

IDO Playbook 2022
Everything we do is focused on enabling fast,

confident decisions and pioneering actions.
We transform our clients with end-to-end data
analytics and AI fueled services, orchestrating
complex business transformations that drive
impact and shape the future.
© 2022 Deloitte LLP. All rights reserved. IDO Playbook 2022 2

Embedding AI, Data and Analytics across an organisation
Do you think about the future?
Do you think about how the way you will live, and work, will change? What role does data, information,
human and artificial intelligence have to play?
Huge disruption has already taken place, the methods in which people and businesses interact, the
types of services being delivered, and the channels through which these services are secured, have
fundamentally changed within a blink of an eye.
Our familiarity with technology and information access in our personal lives has caused a seismic shift
in our expectations as customers, as employees and as citizens.
Every industry and public sector body is facing challenges to the status quo. It’s no longer enough to
experiment with AI and its potential – organisations without a concrete plan on how to scale their use
of AI will face significant threats from new business models and services.
This playbook offers food for thought and aims to help shape your organisation’s data and AI agenda.
It will help you to assess what kind of organisation you want to be, which capabilities and fundamental
building blocks are required, and how to structure the way forward.
Natalie Williams Andy Gauld

Partner, Deloitte Consulting Partner, Deloitte Consulting
IDO Proposition Lead IDO Proposition Lead

IDO Playbook 2022 3
Contents
Introduction Taking the right actions (cont’d)
5 An Insight Driven Organisation 64 Analytics on the edge
68 Secure collaboration
What kind of IDO do you want to become?
20 What is an IDO?
Making the right impact
75 War for talent

Asking the right questions
80 Adopting IDO Culture
31 Vision
87 Distributed Enterprise
36 Value generation
93 Total Experience
38 Insight Driven Performance
Taking the right actions What’s next?
44 DIY Analytics 98 IDO Scaling Lab
51 Digital Ethics and AI 100 Experience Analytics & our global presence
59 Start small, scale fast 103 Contacts

IDO Playbook 2022 4
Introduction
An Insight Driven Organisation

Transformation is more about people than technology
IT’S TIME TO ACT NOW HOW TO SUCCEED?
In a world where disruption is the status quo and technology is advancing at an In a world where information is often incomplete or ambiguous it is imperative
exponential rate, it is critical for organisations across all sectors to make the that organisations collect high quality data, undertake effective analysis and
smartest business choices possible, maximising efficiency and effectiveness. develop and test hypotheses to aid decision making. Leaders must understand
the impact of these decisions, continuing with the hypotheses if successful, or
The global context in which businesses operate today means that competing
failing fast to learn even faster. Organisations which embrace disruption
with industry rivals is more challenging than ever. Many industries contain
through insight will innovate, grow and out-perform their competitors.
players who offer similar products, utilising comparable technologies. Global
competition has reduced location-based advantages and proprietary In order for this to be possible, businesses need to nurture a culture that values
technologies are quickly mimicked. Similarly, the public sector faces increasing insight, curiosity and taking risks. They should not shy away from conflict –
pressure to provide services which deliver tangible impact, via the most efficient having insight means debate will naturally occur on the best way forward. When
means possible. functional business teams begin using insights to debate the best course of
action, it is an indicator of success!
Organisations that will not only survive but thrive in this ever-evolving reality
understand the critical importance of incremental gains. Leaders of the
organisations of the future are constantly searching for new sources of
advantage, differentiation and impact, at a time when they have more data than
ever before. In order to remain relevant, competitive and deliver their purpose,
organisations will need to invest in data, analytics and AI capabilities and take
their organisations on an insight driven journey. Those who lead and innovate
will do so by integrating data and analytics into core business and decision-
making processes, utilising insight at every available opportunity.

IDO Playbook 2022 6
Winning organisations understand innovation goes further than just the technical
ORGANISATIONS THAT CREATE A WINNING

INNOVATE BEYOND THE TECHNICAL
CULTURE ADOPT:
Digitally mature organisations recognise that data & insight is just part of the
tool kit for innovation and growth. Whilst becoming insight driven can help to
promote more technical innovations such as cost savings or greater visibility
over supply chains, organisations that understand innovation stretches beyond
the technical into anything useful to business, are the ones which see the
greatest impact from insights.
Organisations which have placed time and effort on creating a winning culture 1. A detailed and 2. A culture that 3. A challenger
and attracting the best talent through building the ethos of challenging the nuanced prioritises insight mindset and
status quo, or creating distributed teams that are empowered by ownership understanding of over instinct willingness to
the customer disrupt
over their decision making, can lead to innovations far greater than the technical
ones, creating whole new experiences, products & services.
4. Distributed 5. Constant Innovation and 6. Proactive governance

decision-making learning which inspires and and ethical decision-
and co-creation engages talent making

IDO Playbook 2022 7
Leading means integrating strategy, people, process, data & technology
IDO ACCELERATES
Strategy
Becoming an IDO is not about replacing human judgement or experience with technology. It is about augmenting, accelerating,
and enhancing decision making with technology. Insight Driven Organisations don’t entrust data and analytics activities to one
group or team, they manage analytics at an enterprise level with an end-to-end view across strategy, people, process, data and
technology. Organisations who fail to make their transformations stick lack integration across these, sometimes taking a blinkered
view to digital transformation that focuses heavily on data and technology. People
Becoming an IDO means adopting a holistic view to analytics transformations, with five crucial dimensions to an Insight Driven
Organisation:
Strategy - It is essential to identify the key strategic priorities to pinpoint which data & analysis need to be put in place to
attain core organisation objectives Process
People - Insights derived from AI and other forms of big data have now become an organisation’s key currency and
determinant of success. Organisations depend on employees up the entire leadership hierarchy to have the capacity to
comply with the expectations of an IDO-led culture
Data
Process - A successful IDO depends on leaders and employees who communicate, apply, and model well-defined practices
Data - Organisations that use insight driven approaches to decision-making rely on comprehensive research or data
gathering, to build trusted insights that can be used by leadership
Technology
Technology - A robust and secure technology infrastructure is the foundation of a successful IDO

IDO Playbook 2022 8
Human intelligence is accelerated, augmented and enhanced with AI. The potential is limitless.
Let’s make AI our greatest human ally
Automation with Intelligence

Automate data workflows, analytics and models to supplement human decision making
Your business processes with speed and precisions

IDOs
Embrace AI across business processes considering speed and precision as value drivers throughout
IDOs enhance
enhance
decisions with
decisions
with
people and
people and
technology
technology Data & AI with ethics
Work ethically with machine learning models and ensure risk of bias is appropriately dealt with
Collaboration made greater with the machines we invent

Innovate AI & data capabilities that empower workforces to work better together – this is about
AI working hand in hand with humans

IDO Playbook 2022 9
Now AI is everywhere – What’s next?
AI Progress
In our previous edition of the playbook, we told you that AI would become the
new electricity …predicting that
our future is
Artificial intelligence has changed the way we work, the way we communicate, the …and then it all sooner than
happened within a
way we make decisions and the way we interpret the world. most think
matter of ~20
Organisations have leapt at the opportunity to embrace AI – 93% of years…
organisations agree that AI is important to remain competitive in the next 5 years
and 85% of them are increasing their investment in AI in the next fiscal year. …at an
…It took ~45 astonishing
years of almost pace…
no progress to
find traction…
Now we’re telling you AI at scale is next

Organisations now face the task of shifting from AI Experimentation to AI at Scale. This kind of
While most organisations have begun the process, few are fully transformed. linear tend is not
how AI
As companies experience the limitations of big data as a critical enabler of analytics developed
and AI, new approaches known as “small data” and “wide data” are emerging.
Handled correctly, the value that AI generates should pervade the way an Time
organisation operates, beyond affecting only hard metrics such as cost or revenue.
The journey and vision to fully realise the potential of AI at Scale will differ by
1956
1973
1987
2000
Today
Our Future
industry and global region. Projected
That’s where we come in.

IDO Playbook 2022 10
Many organisations have already begun their journeys to becoming insight driven
SCALE MAXIMISES IMPACT
Many organisations today accept that richer insights delivered in a timely

manner will benefit them exponentially, however executing the large scale
transformation programmes required to become an Insight Driven Organisation
is challenging for most.
The transformation required for insights to be utilised at scale, consistently and
continuously, until they are embedded into the very fabric of the organisation is
no mean feat. Many leaders face a common challenge of demonstrating the
value of the investments required to become an IDO.
Some of these investments may take the form of new technical architectures,
which are not utilised to their full potential or recruitment of top talent into Data
& Analytics teams, who fail to gain traction with business functions.
Delivering at scale is therefore a critical component of demonstrating the value
of becoming an Insight Driven Organisation.

An IDO leader will ask themselves key questions when evaluating a new analytical initiative
Market Conditions
To what extent will this investment make us more agile to
Competition respond to changing market conditions?
How will this investment make us more competitive?
How will it help us win and differentiate ourselves Scalable

against competition?
How does the initiative improve our enterprise-
wide analytical capabilities at scale?
Innovation & Growth Data & Technology

How will the investment foster greater Does the right data exist? If not, can we get it or
innovation and growth opportunities? create it?
Is the data timely, consistent, accurate and
complete?
Operating Model Is the technology reliable? Is it cost effective? Can

it be scaled and distributed across my enterprise?
What complementary changes needs to be made in order to take full
Is this the right approach or tool for the right job?
advantage of new capabilities, such as skills, better IT, training, process
re-design or job redesign? Is my operating model fit for purpose?

Organisations continue to search for improvements to become a leading IDO. Where do you stand?
Data driven Insight driven
1 2 3 4 5
Analytically Impaired Localised Analytics Analytical Aspirations Analytical Companies Insight Driven Organisation
No enterprise vision and strategy that Little awareness of analytics vision, Vision for insights developed and Enterprise wide vision and strategy with Mature insights strategy and vision, all leaders
embeds data, analytics and AI. Poorly value and analytics. BI operates communicated, benefits documented, benefits identified. Analytical solutions and employees live and breathe analytical
Strategy integrated systems, ad hoc work and no independently with little structure. existence of siloed analytical strategies, and align to strategy. Senior leaders solutions. Effective, scalable, agile, regularly
innovation created Benefits identified but not measured a decision taken as to the model through interested in data driven culture challenged and optimised
which analytics is delivered
Few skills attached to specific functions, Isolated analysts, unmanaged mix of Analysts recognised as key talent in key Highly capable analysts recruited. Professional analysts and citizen data scientists.
no active recruitment for data, analytics skills, local recruitment taking place. areas. Basic capability and ad hoc training Upskilling plans in place. Executive Data and analytics is a democratised skill all can
People and AI. Insights not supported, no No roadmap to support change and provided. Immature knowledge champion spearheading analytics contribute to. Clear career paths for employees
knowledge management capability resistant culture management framework exists transformation
Minimal automation. Most data Some effort to automate regular Considerable effort to automate regular Analytics activity linked to business Formal SLAs, governance and analytics with key
processes manual or unchanged. Data processes. Some business partnering processes. Formal SLAs, comms and objectives and prioritised. 90% solution goals and objectives defined and
Process not used in decision making, no SLAs or activities and informal SLAs. Sporadic governance structure but not fully processes automated. Analytics teams communicated. Data key for decision making
governance for analytics data use for decision making. Analytics integrated. Op model limited by legacy IT, work closely with business to change across all functions. Process automation a core
op model not agile processes and data quality existing processes, formal SLAs in place functionality
Inconsistent, poor quality, Some metadata management exists, Key data domains identified, central Integrated, accurate, common data in Search for new data integrations to enhance
unstandardised; siloed. Limited no senior executive discussions, data is repositories created. Sources mapped and central repositories. Information model decision making, structured/unstructured
integration. Difficult to access. No accessible, privacy and security polices recorded, quality is acceptable but allows quick and easy access to data (wide) data models in place, highly accessible
Data ethics, security and privacy policies in defined on project level, limited data inconsistent. Insights developed using sources. Integrated data from multiple at scale, fast querying, master data, enterprise
place collection and quality models, privacy and security polices, ethics sources. Single data definitions, quality wide security
is proven, strong focus on ethics
Analytics production environment non- Limited analytics production Production environment, advanced data Cloud based tech, sandbox testing Sophisticated analytical architecture, AI tech,
existent, no sandbox environment or environment with isolated usage of storage, real time processes, ETL and integrated into IT strategy, advanced BI prescriptive, sandbox environments,
Technology tools testing and solutions. Manual Ad static dashboards. New technologies visualisation, test environment, pre-built dashboards, stable technology and democratised BI tools, predictive technology,
hoc reports. Limited BI. Little cloud adopted on ad-hoc basis. BI accessible reports and dashboards manually updated. optimised, cloud solutions for advanced cloud analytical techniques, scalable and auto-
awareness but driven by complex processes Cloud solutions deployed analytical modelling optimised around demand

Business improvement can be simplified to five value pillars developed by becoming an IDO
Enhance efficiencies Provide superior Provide a future- Impact beyond the Provide greater data
across the business customer service and proofed data ecosystem bottom-line quality & transparency
satisfaction using data able to deliver analytics improving risk
driven approaches which enrich insights management and
reducing margins of error
What does this mean in practice ?
• An enhanced bottom line • Superior customer service • A comprehensive data eco- • Focusing on delivering • Working to increase data
and higher degree of and satisfaction, as data system that provides all change outside of the quality drives increased trust
efficiency, as resources are provides a deeper descriptive and explanatory technical is a key part of of reporting right up the
utilised within the scope of understanding of target analytics the business needs, becoming an IDO. As a C- leadership chain, which
comprehensive, well- audiences, which promotes when it needs it suite this translates as a delivers data driven decision-
planned initiatives customised, relevant and greater ability to outwardly making right to the top
• The same eco-system is
highly engaging consumer improve the impact your
• Senior management have future-proofed and capable • Less time spent on
experiences organisation has on society
the data which provides of delivering predictive and duplication and reworking of
as well as your ability to
visibility on people, • The greater understanding prescriptive analytics where datasets across the
inwardly to retain and attract
processes & performance of the customer allows C- required, to enrich the organisation enhancing
new talent
across the business allowing suite executives to have the insights of your organisation productivity, reducing risks
for the identification of relevant data to decide and driving a data-led
savings and improvements where to start, stop, and mindset
from a board level continue within the business

Business improvement can be simplified to five value pillars developed by becoming an IDO
Enhance efficiencies Provide superior Provide a future- Impact beyond the Provide greater data
across the business customer service and proofed data ecosystem bottom-line quality & transparency
satisfaction using data able to deliver analytics improving risk
driven approaches which enrich insights management and
reducing margins of error
How we deliver
BT Legal Lawn Tennis Association Department of Health & Scope Retail Estate Network Rail
(LTA) Social Care
Deloitte in partnership with BT LTA was able to connect with COVID-19 Response solution Scope wanted to positively Previously with the introduction
used AI for fast and accurate over 1 Million fans directly by resulted in an increase in tests disrupt the charity sector. In of a new train timetable caused
analysis of 4,500 documents to 2022, directly adding an extra returned within 24 hours from partnership with Deloitte hundreds of cancelations.
identify and categorise key 100,000 users through 30% to 75% within 1 month of Scope developed an Deloitte and Network Rail
risks, helping achieve a 50% revolutionising LTA’s online the first InSightIQ dashboard optimisation & innovative worked together to build a
time saving while also presence with personalisation being delivered community impact virtual way to test new
delivering a more exhaustive assessment. The novel timetables saving the 22,000
review of the documents approach resulted in the trains carrying 4,000,000
estimated positive impact for passengers a day from
at least 10,000 disabled people disruption

Becoming an IDO is not achieved without overcoming barriers along the way
Challenging the status quo

When embarking on an IDO journey, it is important not to underestimate the potential for resistance and the levels of education and change which need to take place
across the organisation.
The bigger and more mature the organisation, the more Confidence in data is low due to inconsistent
Cultural Change difficult it is to drive a cultural change or Data definitions. Reluctance to share data and inability
transformation. to get timely access to it.
Customisation Re-inventing the wheel rather than exploring the

Inefficient Delivery Not having a strong operating model where centralised leading to unnecessary market for off-the-shelf elements can increase
and decentralised resources work seamlessly together. complexity complexity as well as time to market.
Analytics is developed in silos and effort is

Inaccurate Metrics Over simplistic models, overconfident analysts, ill- Siloed
duplicated. It lacks implementation vision and
and Models defined outcomes/assumptions lead to incorrect results. Implementation
strategy for enterprise-wide integration.
Image is ‘Techy’, complex, and related to math and Is often stuck in ROI discussions, change inertia,
Technical
statistics, and hence difficult to comprehend or thought Buy in scepticism, fear of being challenged, and under
Perception
to be IT-only. cost considerations.
There is a large supply gap of data analyst and data Organisations are distracted by the hype and are
Tackling
Talent Crunch scientist talent, organisations are shifting towards hiring confused by what Big Data, AI and Robotics really
the Jargon
talent who generate insights - not just number crunchers. means to them and how best to apply them.

What does success look like for an IDO?
1 2 3
Understand the
Utilise accessible impacts of key
insights at scale, decisions, with the
Have strong, visible
consistently and ability to integrate
leadership within the
continuously, until across processes and
organisation for Data,
they are embedded functions to predict
Analytics and AI
into the very fabric outcomes,
of the organisation understanding: what
will happen?
4 5 6 7
Operate in an Agile
Test hypotheses, Understand their and Reactive
Can clearly
learn from them and customers or users manner, embracing
demonstrate the
use insights to solve intimately, blending disruption with the
value insight is
problems and internal and external ability to
unlocking across the
identify new growth data sets to create continuously
organisation
and value areas richer insights optimise operational
processes

So, how do you adapt?
The IDO Playbook provides a flavour to becoming an IDO and how organisations can face some of the
common barriers to do so
START WITH THE BASICS - THE IDO PLAYBOOK ORGANISATIONS ARE TALKING!
1. Understand what it means to be an “IDO” of our global IDO survey* respondents have told us developing improved data,
50% analytics & AI strategies as well as driving support for digital transformation –
the “distributed enterprise” – are the top two areas of investments
2. Examine whether you are :
a) Asking the right questions? of respondents express defining value and prioritising projects as a strong
52%
barrier when trying to generate analytical insight
- Do you have an insight-fuelled vision that generates value?
b) Are you taking the right actions?
state they do not have the right data available or accessible for analysis within
- Are you building trust and digital ethics into your AI stack? 57%
their organisations, closely followed by the challenge of managing data quality
c) Are you making the right impact?
- Is your culture set up to become data driven at scale? expressed the need for AI skills that they cannot source – over 10% increase on
55%
last year. This was also reflected for data science skills
3. Identify the key operating model building blocks your organisation needs to
focus on to become an IDO (Strategy, People, Process, Data, & Technology) told us they do not use AI in any of their business functions predominantly due
60%
to a lack of priority and skillsets
4. Check out our IDO Scaling Labs propositions in order to accelerate your
transformation towards becoming an IDO, and put theory into practice. of respondents stated that there is “no visible and active leadership” in the field
20%
of AI, driven by a lack of clear accountability towards AI within organisations

*Global IDO Survey 263 total respondents across 17 countries IDO Playbook 2022 18
What kind of IDO
does your organisation
want to become?

What is an IDO?
An organisation driven by Intelligence, Insight and Information, in order to be Innovative at
an Industrial scale
An IDO is an organisation that has a ‘data first’ mindset, seeking to infuse AI, analytics and robust data into all parts of the
organisation, especially within their decision-making processes. They do not view this as a project with a start or end date, but rather, will
continuously ask how they can systematically leverage data to improve or reinvent processes to enhance business efficiencies. IDO’s
see AI and analytics as core capabilities across their organisation, to provide improved products and services, to support business
development and process optimisation; to tackle their most complex business problems; and to address growing analytical trends..
The I in IDO now covers the following 5 areas:
Information Insight Intelligence Innovation Industrialisation
Robust, secure and Strong analytical Leveraging intelligent A culture of ideation Enterprise-wide
reliable data storage, capabilities, leveraging automation, cognitive and knowledge collaboration,
good quality data with descriptive, predictive analytics and AI to sharing, with a integrated
strong ETL, reporting and prescriptive supplement and strategy that leverages architectures into core
and BI capabilities analysis. augment human the latest technologies systems, process, and
leveraging the right decision-making. and stays ahead of the behaviours with
tools, platforms, competition. robust processes and
architecture. workflows.

Information Driven Organisation
Robust, secure and reliable data storage, good quality data with strong ETL, reporting and BI capabilities
leveraging the right tools, platforms, architecture
What does good look like?
Strong foundational data capabilities may seem like the bread-and-butter of a successful organisation, but in
today’s landscape this is getting increasingly complex.
The key features are:
✓ Coherent application architecture, with a roadmap in place to decommission legacy/duplicate systems
✓ Accessible internal and external data sources, structured for business consumption through reporting and BI
✓ BAU operations which are reducing manual effort where possible, automating data wrangling and ETL processes
✓ Data quality addressed at source, with validation and on-going monitoring
✓ A focus on securing enterprise data assets from any unauthorised infringement
✓ Everyone understands their role in keeping the organisation compliant with regulation
Are you asking the right Are you taking the right Are you making the right
questions? actions? impact?
• Does the information available, and the way it is • Does your technology stack to support your • Do you have the right technical talent and
used, contribute to your strategic objectives? business objectives and aspirations? training to support analytics capabilities?
• Are you investing in data as an asset? • Does your data governance cover ethical and • Is your organisation empowered to make
• Are you challenging the information paradigm? privacy related concerns? decisions through the use of trusted
information?

Insight Driven Organisation
Trust in insight, leveraging descriptive, predictive and prescriptive analysis and driving data-driven
decision making
Driving business change through insights requires more than data and technology. Transforming analytics to
streamline decision making across all functions requires a cultural shift. The key features are:
✓ A documented, published and embedded insights strategy that is adopted by the business and IT communities
✓ Leadership-level support of the agenda, understanding, prioritising and tracking the value of insight initiatives
✓ An ecosystem of talent with the right blend of business and technical skills
✓ Aggregating and combining data from broad sources into meaningful content and new ideas
✓ A repeatable insight process to test and industrialise analytics
✓ Human-centered design to deliver visual and intuitive insights
✓ Redesigning processes, operations and products based on the insights generated
• Are benefits clearly identified, recognised across • Is the most fitting type of analysis being applied, • Is there a culture of using data to answer
the organisation and have business owners? for example, leveraging edge analytics or wide business questions and inform decision making?
• Do insight capabilities exist across your data where required? • Are insights delivered to enhance the brand
organisation and are they used effectively? • Are insights being used to re-engineer, augment experience for your organisation?
and improve business processes?

Intelligence Driven Organisation
Intelligent automation, cognitive analytics and AI to supplement human decision-making
Existing ways of working are made more efficient, focusing human input on making decisions and acting on them,
rather than collecting and analysing data. Emerging technologies can automate traditionally manual processes and
begin to learn and improve based on feedback. The key features are:
✓ Proof-of-concept ideas leveraging emerging technologies are explored as part of ‘business-as-usual’ (e.g. scripted
task bots, natural language processing, speech and image recognition, machine learning enabled analytics)
✓ Effective proof-of-concepts are being piloted and productionised to improve operational efficiency or develop new
products and services
✓ Ethical governance forums test the long-term implications of new solutions/ways of working
✓ Tools and technology are well integrated – i.e. there is minimal FTE to maintain systems
• Does leadership understand the opportunities • Are your processes intelligently optimised? • Do you have the right blend of business and
offered through emerging technologies? • Are ethics and privacy concerns addressed technical skills to identify, develop and embed
• How do you see intelligence as a value generator within your processes? intelligent solutions?
within your business? • Are you testing new AI capabilities? • Do your intelligent solutions cater for dispersed
• How do leaders in your industry leverage data as employees, customers and assets?
a competitive differentiator?
Innovation Driven Organisation
A culture of ideation and knowledge sharing, with a strategy that leverages the latest technologies
and stays ahead of the competition
A culture of continuous innovation is embedded into the organisation, enabling it to adapt to market changes. The team
rigorously challenges status quo and has an established feedback loop for transformation initiatives. The key features are:
✓ Innovation is integrated into regular business activities and consistently is seen to update products and services
✓ Employees are empowered and provided training on the latest technologies
✓ There is space for innovation within teams, whether this is virtual or actual, to build a community around new ideas
✓ There are processes in place for new ideas to be tried, tested and productionised
✓ The organisation has adopted a ‘fail-fast’ agility
✓ Innovation is seen as a priority for all, not just specific innovation or incubation teams
• Is there leadership investment for innovation • Do you have an ideation process? • Are you using new insights to continuously
initiatives? • Do you have a sandbox environment for rapid enhance the products and services you offer?
• Is your vision well understood to support development? • How do you ensure lead time remains
valuable innovation in the right areas? competitive?

Industrialisation Driven Organisation
Enterprise-wide collaboration, integrated architectures into core systems, process, and behaviours with
robust processes and workflows
A culture of data literacy is present throughout the organisation, teams are working collaboratively to understand and
deliver value from data. The key features are:
✓ Knowledge management forms a key part of the insights culture, sharing of algorithms, visualisation and reporting
best practices and troubleshooting is common across the organisation
✓ The business, analytics, data and IT teams work together, especially in times of high demand or transformation
✓ Analytics solutions exist within an operating model which incorporates their governance, alongside lifecycle
management
✓ There are continuous quality assurance processes, checking-in on ethical bias and algorithm accuracy, adopting
continuous self-learning and making adjustments
✓ Service Level Agreements are in place with the customers of analytical solutions
✓ Change management is in place to embed any new solutions and ways of working, and reap the rewards
✓ There is continuous benefit tracking for each analytics solution
• Are you making better decisions by joining the • Do you have the appropriate design authorities • Are there sufficient resources with the right skills
dots across business siloes? in place to make organisation wide changes? to deliver at scale?
• Are you enriching the depths of analysis by • Are you integrating ethics, security and support • Do you have strong relationships across the
learning and sharing across teams? into your development capabilities? business to monitor, understand and adapt to
• Are you democratising analytical applications? business need?

What is an IDO?
An IDO builds a culture of asking the right questions and constructs an ecosystem of data,
digital, cognitive and innovation capabilities to support an analytical culture
The building blocks

required to become an IDO Vision
Choice
Value Drivers
Stakeholder
Strategy &
Structuring
&
Management
Innovation
A structured approach is essential for Purpose Business Case
achieving both pace and quality in
creation and adoption of analytics
solutions at scale. Culture
Organisation
People Leadership Talent & Collaboration
An IDO’s capability goes far beyond Design
Change
IDO Operating Model

technology – it considers strategic
alignment, talent and leadership,
business processes and the entire
Delivery
information lifecycle and systems Demand & Process Benefits
Process & Governance
associated. Prioritisation
Operations
Optimisation Realisation
In its entirety, the IDO building blocks

helps define your organisation’s
target operating model Data Quality Trust
Information Data as an Regulation &
Data &
Modelling Asset
&
Compliance
Management Digital Ethics
Cloud & Sandboxing Security,

Technology AI
Technology Platform
Blueprint Enablement
& Reliability &
Architecture Industrialising Continuity

What is an IDO?
Considering AI with a wider lens than just technology
Asking the right questions Taking the right actions Making the right impact
To fully embark on your organisation’s analytics journey, you must In order to improve the analytics maturity of your organisation, Success can only be achieved when your organisation is
first build a strong strategic direction and an understanding of the you must select and invoke the right data and analytics initiatives supported by the right capabilities. Education and culture are key,
value proposition in order to support this journey as well as managing ever-changing business and customer models
Organisations can unlock value at pace by Creating an ecosystem of analytical talent with
Ensuring the AI vision aligns to and DIY Data combining capabilities from different cloud The War for the right balance of business and technical
Vision supports corporate goals and objectives. Analytics vendors to build their technology stack Talent skills – attract more data scientists and AI
whilst democratising analytics. experts into your organization.
Designing and enabling the appropriate

Understanding the value of AI and Digital Ethics can be embedded with AI to
Value Digital Ethics behaviours to ensure IDO is embedded
prioritising initiatives appropriately to ensure insights are trustworthy and IDO Culture within the organisation and continues to
Generation support a business case. and AI technology aligns with company values.
deliver long term value.
Insight Evaluating how businesses should be using Shifting towards less data hungry models Combine MX, CX, EX and UX to create an all-round
insight to manage their business Start Small, or combined structured and non- Total exceptional experience. Architected correctly,
Driven performance and how to unlock value Scale Fast structured data to overcome limitations of Experience organisations cam unlock powerful insights from
Performance through effective decision making. large data sets and optimise costs. total experience technology.
Capture more insights by tapping into How can we organise to deliver value in a
Analytics on assets distributed outside your cloud Distributed dispersed business model with employees
the Edge environment and reduce analytical latency Enterprise and customers in varying locations? What
by performing analytics locally. technology should you be investing in?
Share and access more data whilst preserving

Secure privacy and security – tap into datasets that
Collaboration traditionally are difficult to access due to
regulatory concerns and high sensitivity.

What is an IDO?
How do you address key organisational challenges as an IDO?
Any organisation, regardless of size, industry or focus, will encounter hurdles on their journey to embedding data and analytics into their business.
The IDO Playbook can help reshape your thinking as you address these challenges while providing some insight into the art of the possible.
Asking the Right Questions Taking the Right Actions Making the Right Impact
Value Insight Driven DIY Data Digital Ethics Start Small, Analytics on Secure Distributed Total
Theme Problem Statement Vision
Generation Performance Analytics & AI Scale Fast the Edge Collaboration
War for Talent IDO Culture
Enterprise Experience
“I am having difficulty finding the right

skills internally and / or hiring externally
Talent Crunch
that are needed to embed advanced
analytics into my organisation"
“The data quality in my organisation is

Data Quality poor, and thus, nobody is using it as
confidence in the data is low”
“I can’t access the data I need in a timely

Data
manner / at all, and my colleagues are
Democratisation
reluctant to share data with me”
“I want to be confident that our data

governance initiatives adequately address
Governance
all ethical, privacy and biases-related
concerns”
”We don’t have a culture of using analytics

across the organisation, and adoption of
Culture
data is slow to non-existant. Instead
people are fearful of change”
“Production of analytics is dispersed across

Siloed teams the organisation and sometimes
and insights duplicated. Processes aren’t standardised
and tend to be slow and manual”

What is an IDO? (cont.)
How do you address key organisational challenges as an IDO?
Any organisation, regardless of size, industry or focus, will encounter hurdles on their journey to embedding data and analytics into their business.
The IDO Playbook can help reshape your thinking as you address these challenges while providing some insight into the art of the possible.
Asking the Right Questions Taking the Right Actions Making the Right Impact
Value Insight Driven DIY Data Digital Ethics Start Small, Analytics on Secure Distributed Total
Theme Problem Statement Vision
Generation Performance Analytics & AI Scale Fast the Edge Collaboration
War for Talent IDO Culture
Enterprise Experience
“I don’t think my analytics team

Conflicting understand what we are trying to achieve
priorities as a business, and my various business
units have conflicting data priorities”
“There is no one in the business

Lack of
championing analytics, no data ownership,
leadership &
and no accountability around the
accountability
production and use of analytics”
Customer ”How can I use analytics to enhance the

Experience customer experience?”
“How can I utilize analytics to run my

Benefits organisation more efficiently and
realisation effectively, and how do I measure these
benefits?”
“How can I remain innovative and

Cost vs
competitive while remaining cost
innovation
effective?”
”There’s too much jargon in analytics to

Poor
understand its application to my business,
understanding
and it’s hard to differentiate it from the IT
of analytics
function”
"How do I keep up with the market in

Scalability & terms of methods and technologies in
Agility such a fast-moving world, and ensure
these are scalable and agile?"

Asking the right questions

Vision
What does becoming an IDO mean to our business?
THE VISION: MORE THAN JUST DATA AND TECHNOLOGY

Organisations often fail to realise the true potential of data and AI due to the
assumption that successful AI and data delivery is exclusively IT driven; only
considering the gathering of data from source systems into an analytics or
customer-facing solution. This can lead to a disconnect between the decision-
making process, business value, and analytical insight. An AI and data vision
needs to support the overarching business strategy, embedding insight into
the underlying drivers which move an organisation towards its goals.
WHAT DOES SUCCESS LOOK LIKE?
To deliver successful analytical insights, organisations need to develop
solutions that are underpinned by the right capabilities; organised in an
appropriate operating model; governed in a structured way; and supported
by people with both technical and business literacy who can translate
knowledge into clear business value impacting decision-making within the
organisation. Technology is used as an enabler (but not a driver) of
this process.
Developing a clear vision and narrative for the IDO journey will help
organisations align behind a set of common objectives and invest only in
capabilities which support these strategic goals.
Failing to articulate an AI and Data Vision is a common reason for failing to
make analytics ‘stick’ in organisations.

Vision
Future proofing your IDO vision

A clear vision helps to drive towards a common goal. This means avoiding the traditional temptations of a 3-5 year transformation programme is necessary to ensure
responsiveness.
Organisations need a malleable AI and Data vision which is flexible and durable as the organisation learns more about the disruptive environment it is operating in. This
flexibility will allow insights to be adopted along the way. Other considerations and constraints when future proofing the AI and Data vision include:
A demanding, expensive workforce who find traditional structures increasingly prohibitive.
Policy-makers and regulators are reacting inconsistently to data privacy and security issues.
Markets and ecosystems reluctant to relinquish traditional means of quantifying success.
Technology breakthroughs happening faster than ever before. Data explosion leading to a lack of innovation.
Disruptive, agile start-up organisations which use digital means to revolutionise service offerings.

Vision
New revenue streams Better together?
Effective use of an organisation’s data, including the Looking at data opportunities as a multi-stage value chain, companies can decide at which stage they
incorporation of AI where appropriate, can open up want to operate, depending on maturity levels.
opportunities for new business models to broaden and
increase revenue streams. Organisations can begin to
assess the type of business model which might best
apply by answering the following 4 questions:
1. Which other companies in the industry ecosystem
apply data-driven business models successfully and
what value do they derive?
TOP TIPS
2. What is the approach to implement the business
model and how significant is the required change?
• In a low-level maturity business model, multiple stages are owned by a single
3. What methods exist to assess the monetary value of company, because no other players exist
data assets?
• When competition increases and the data value chain matures, stages tend to be
4. What constraints and regulations with respect to owned by separate, specialised companies
data privacy apply to these business models?
• Knowledge about industry maturity, which companies work together to enable a
Whilst the latest trend of data monetisation can look
data value chain, and the company’s own capabilities, can help to decide which
attractive, it is not suitable for many organisations –
and they could end up selling their most valuable asset. business model to pursue
By understanding which business model suits the
organisation best, a strategy to generate revenue and
efficiencies from data can be defined that is tailored to
the company.

Vision
Having effective use of your data has delivered clear value and has helped to establish market leaders. A failure to address the increasing prevalence and effectiveness of
technology can lead to your organisation falling behind. Possible organisational gains from putting AI and Data at the heart of your business can be split in to five
categories:
Intelligent Cognitive Insights Transformed Fueled Innovation Fortified Trust

Automation Engagements
Automate the “last mile” of Improve understanding and Change the way people Redefine “where to play?” Secure the franchise from
automation by removing decision making through interact with technology- and “how to win?” by risks such as fraud and cyber,
humans from low value and analytics that are more allowing businesses to enabling creation of new improve quality and
often repetitive activities proactive, predicative, and engage on human terms products, markets and consistency, and enable
(often in service of machines) able to see patterns in rather than forcing humans business models greater transparency to
increasingly complex sources to engage on machine terms enhance brand trust
Gigster automates software By equipping trains with AI- Amazon Go enables retail OakNorth’s “ACORN Machine” HP’s Aruba Introspect analyzes
development project powered sensors that leverage customers to enter a store, analyzes alternative data to network traffic to identify
management by quickly advanced analytics, Trenitalia automatically add items to quickly originate bespoke loans subtle anomalies using
estimating the work required, streamlined its unplanned their virtual cart, and leave for small and medium sized behavior-based attack
organizing experts, and maintenance and increased store without needing businesses, growing its loan detection to proactively
adapting workflows productivity to check out book to ~$1B in two years quarantine affected devices
in real time

Vision
Often it is difficult to see how a vision can have the impact and work in practice in transforming your organisation to put analytics at the heart of your business. Here we
can see that the articulation of a clear vision will have a cascade of effects on your organisation right down to the people, process and ways of working.
Set Vision & Objectives Define Mission Priorities Ensure Operational Alignment
What are our What

analytics goals What
Where will How will management
capabilities
and we focus? we succeed? systems will
will we need?
aspirations? we need?
Defining your vision to become Establishing the focus, as Winning in the areas you have Understanding the capabilities, The final layer to executing on a
an IDO organisation will allow determined by your IDO based chosen to focus on is essential you need to facilitate your vision is to understand the
you to set goals which deliver a goals will enable you to narrow to deliver on your goals and thus success and in turn your IDO granular impact on the way your
clear barometer on success of the playing field and understand vision. Vision means your analytics organisation works.
your transformation. where you can deliver the best Winning can be split in to three transformation will go further These are questions like:
Your vision and goals will allow business outcomes. main areas which should be than just technology. determining the operating
you to understand the level of Asking key questions like where addressed within the IDO Considering not just the model that best fits your
ambition and investment we will deploy our analytics, and framework: technology required, but the capabilities; how can you create
required, as well as considering who exactly will benefit will help • Winning technically processes and people required an IDO culture; and how can you
often forgotten questions on to determine where is best to to deliver for your measure success in practice are
• Winning organisationally
ethics, privacy and transparency focus in order to reach your transformation is key. key to becoming an IDO and
around data. goals. • Managing risks realising your IDO vision.
This will make analytics “stick” in
your organisation.
Value generation
How can we understand and measure success?
A successful organisation leverages analytics across the business to gain value by:
What does success look like?
Organisations often find it difficult to measure and Identifying business process improvements
demonstrate the value of AI and data analytics
projects back to the business. This inability to show
ROI stems from one of 3 root causes:
Understanding and improving customer experience
The project is not aligned to business

1 needs or designed to answer a Guiding company strategy
specific business question
Deriving performance measurement / ROI metrics
Despite a well-executed project, the Informing marketing and communications strategies

2 business has not adopted the
insights or the solution in BAU
Creating or modifying products and services
Monitoring the competitive landscape

Unrealistic expectations were set in
3 terms of results, timelines, outputs
or user-experience
Targeting sustainability metrics and goals

Value generation
How can we understand and measure success?
High level assessment Short list and prioritisation Granular understanding
Demand prioritisation
Each AI and data analytics project should be assessed on the value it delivers (strategic and financial) and the risk involved. To enable the business value to be consistently
measured across all projects, they should be assessed against each value driver to establish a relative score for the perceived benefit to be delivered to the organisation.
When assessing projects, the impact on people and organisation, the technology as well as the data and governance should be taken into consideration since change can create
significant value as well as risk.
The prioritisation process
Often businesses have several AI and data analytics projects sponsored by

different departments competing for the same delivery resource. This cross Longer-term priority Immediate Priority
departmental demand makes it difficult to effectively prioritise projects which are High value but will require High value and likely to
attempting to deliver both direct or indirect benefits to the business. significant investment in demonstrate that value within a
An objective prioritisation process needs to be put in place, with cross capability build to deliver reasonable timeframe
Value
departmental representation, to identify those projects that deliver the greatest
value with the least risk and achieve business input and ‘buy-in’ on their execution.
Lowest priority For consideration
A prioritisation process has 3 key outcomes:
High effort with too low value Reasonable return on investment
• Measure the value and risk for each potential project
to be prioritised and should be prioritised
• Cross departmental agreement of project prioritisation
• Clear list of projects for the delivery team to progress
Ability to capture
Insight Driven Performance
Making informed choices reliably and repeatedly using the industrialised generation of insights
within an organisation
Effective decision-making is fundamental to realising success for any organisation. Each decision made – big or small, strategic or operational – is a nudge on
the organisation’s steering wheel. How well an organisation selects and co-ordinates those nudges will determine its direction of travel, how long or bumpy
its journey is and whether or not it actually gets where it wants to go.
Insight Driven Performance connects Strategy to Execution

in an iterative and self-improving system
Learnings from the operational level are fed into the strategic 1
decision-making choices of the organisation
Insight
Driven
Performance
3 2
An iterative approach – from insight generation through Prioritised areas of performance and clear
action, tracking and learning – is established to help the accountabilities are embedded in the organisation’s
organisation pursue strategy and learn-fast at an performance management approach to support the
operational level adoption of strategy

Insight Driven Performance requires the organisation to not only generate insights, but to take
the informed actions necessary to turn those insights into enterprise value
Data The facts and figures available to be reviewed
Dialogue Discussion and analysis of the data in context
Accurate and deep understanding of the

Insight Driven INSIGHT true nature of the situation
Performance
Informed action Insight Driven choices put into action
Positive outcomes for the business as

Enterprise value a whole, be it improved sales, profitability or
reputation

Thriving in a dynamic environment: The insight-driven qualities of leading organisations
Disruptors to all organisations
Demands of the
Emerging Technologies Exponential Data Growth System-Shocks
Connected World
Qualities of organisations primed to succeed
Collaboration – within an organisation and within its New technologies can make the generation of insight more
business ecosystem – unlocks both innovation and Technology rapid, reliable and repeatable across the whole of an
Collaborative
opportunity and is a source of competitive advantage for enabled organisation providing them with the tools to drive
those organisations who can get it right consistently improving performance
As successes and mis-steps come and go, it is an Agility and adaptability can’t insulate against systemic-
Learning- organisation’s ability to learn that will determine the shocks, but they are qualities that allow organisations to
Agile
oriented relative balance of the two – and which of the two an adjust to those shocks faster and more effectively than
organisation is destined to repeat their competitors
Organisations who acknowledge the importance of data as

the raw material of insight generation and put it practically
Data-led
and culturally at the heart of their decision-making will
consistently outperform those organisations who don’t

Insight Driven Performance depends on eight core components to realise enterprise value
and iteratively improve
Rather than being overwhelmed into inaction, those who succeed will be
those who understand the value of improvement and understand that the
varied components represent ample ways to make a difference
Steering Planning
Business Choices are where the organisation makes its

up-front and ongoing decisions as they conduct their
performance management
Ownership & Insight
Oversight Solutions
Enablers are the components which provide the Talent & Rituals &
organisation with the capabilities to generate insights Teaming Routines
Next Gen End-to-End

Tech Data
The maturity and effectiveness of an organisation will vary across the
different components, but seeking to understand, assess and selectively
improve the maturity and effectiveness of the different components is a
vital step for any organisation seeking to optimise its decision-making

Taking the right actions

Taking the right actions
Following business disruptions of the pandemic, leading organisations are accelerating analytical
capabilities at scale – don’t get left behind!
Depending on whether you are an established organisation, looking to expand into new hot topics or experiment with cutting edge concepts and
technologies, the below provides some examples of what organisations are talking about. In this section, we deep dive into some of these topics.
Establishing Expanding Experimenting

The rise of
Relational
Data the Data Cloud Smart Cognitive Scalable Synthetic Composable Privacy Enhancing Decision
Databases
Visualisation Scientist Crowdsourcing Analytics Robots Analytics AI Data Data Analytics Computation Intelligence
Business Big Data Managed Text Augmented Distributed From Big to Edge Digital ML Ops Total
Intelligence Analytics Analytics Analytics Analytics Enterprise Small & Wide Analytics Ethics & & Model Ops Experience
Data Smarter AI

DIY Analytics
Shifting to a composable D&A (data & analytics) architecture enables businesses to build applications by
interconnecting analytical capabilities via APIs so end users are provided with a catalogue of business
intelligence applications to choose from based on use case
“Composable Analytics = Embedded + Low/No-Code + Consumerisation”
Composable D&A enabled by cloud computing

User
User centric
We are seeing more and more open D&A applications that enable cross centric
communications with other applications via a cloud-based architecture. In
fact, by 2023, Gartner suggests 60% of organisations will compose Agility
Agility
Low code
Low
components from three or more vendor solutions to build comprehensive
analytics applications.
The cloud is fast becoming an ideal place to achieve advanced analytics DIY
capabilities, such as composable D&A, due to its scalability, reduced Analytics
operational effort required and open technology that sits behind it. It is
however unrealistic to implement a full analytics stack at once and many Collabora
companies will do so in stages. Businesses with complex and Flexibility
Flexibility Collaborative
tive
unpredictable needs can adopt composable D&A to bring analytics closer
to the end user or business process. This improves decision making,
creates analytical flexibility and allows more advanced analytic capabilities Innovativ
Innovative
to be hand picked by users for a business process. e

DIY Analytics
Composable analytics drives innovation
1 2 3
Improved APIs democratise More Open Data Models brings Data Governance:
powerful analytics capabilities together Analytics catalogue
Growing technology partnerships between Enables analytical tools to coexist with other Analytics catalogues are emerging through
vendors enables tools to communicate better tools while running on the same data model. adoption of container-based services and
with each other making it possible to package With convergence of ABI platforms and DS business intelligence. Provides user friendly
them within applications – for example – (Data Science) or ML (Machine Learning), interfaces to access reports, models,
deploy a predictive model from vendor X into increased openness of the data model dashboards from a diverse range of
Tableau dashboard to enable more users to structure that sits behind is important to applications. This enables collaboration and
use interactive predictive analytics provide more opportunity for composition drives increased usage of an organisation’s
analytical portfolio
4 5 6
Bridging data science components Transitioning from integration Encapsulating packaged
into analytic applications to innovation capabilities with a composable
mindset
Taking augmented analytics to the next level. Instead of building new dashboards or Lots of analytical tools is great but this cause
More connecting mechanisms will be made embedding new analytical tools, innovate complications when they are not organised
available to bridge components from different solutions to business problems by combining properly into business understandable analytic
analytic domains without rearchitecting the existing analytical capabilities blocks. Composable analytics combines and
existing production process in each platform optimises these blocks as packaged
capabilities from which users can access the
analytical power of multiple blocks
Source: Gartner
DIY Analytics
Leaders should consider Composable D&A techniques in order to…
PROVISION MOST The culmination of intelligent applications

ACTIONABLE working together enable organisations to
INSIGHTS provision most actionable insights possible
Building using low to no code
DEMOCRATISE
tools involves more employees in
the build of analytical solutions
ANALYTICS
Build new capabilities from existing

REDUCE COSTS applications removing the need to invest in
solutions from scratch
Development involves business BRING USERS

users throughout to deliver the CLOSER TO
best solutions DEVELOPERS
BE CONSUMER Build bespoke solutions that are shaped to

DRIVEN solve specific consumer problems
Re-usability of analytical
capabilities through connections RAPIDLY
enables business to build new INNOVATE
applications at pace

DIY Analytics
Data and analytics leaders should explore piloting composable D&A in the cloud and enhance
analytical capabilities by establishing composability within existing analytical portfolios
Impacts Recommendations
• Extend analytics capabilities by establishing composability within

Microservice based analytics, BI, ML and data science platforms existing analytics porfolios
with improved APIs enables assembly of analytics capabilities • Leverage composable analytics to drive innovation by
across platforms incorporating advanced data science and machine learning
content into analytics applications
Low and no code capabilities enable more customer focused • Explore opportunities to add analytics capabilities to applications
analytic application design, improves collaboration with by building a joint team of application developers and business
business user and developer analysts
Popularity of cloud-based marketplaces drives organisations • Pilot composable analytics in the cloud, establishing an analytics
to explore innovative analytic capabilities. Adds value to marketplace to drive and support collaboration across business
cloud-based applications users, developers and citizen data scientists

DIY Analytics
A composable data architecture is built ground up, combining the best of D&A capabilities via
open APIs into low code platforms from which services are accessed via bespoke user interfaces
User Journey: solving an

Leveraging hybrid or Compose user services from a vendor agnostic flexible Composable D&A Adaptation
multi-cloud analytics landscape of best-in-class tools, flex to user need insight-based problem
A multi vendor landscape enables a

Identify business
more wholistic approach and a wider
problem
Vendor Vendor
perspective
Reporting, OLAP, Dashboards
1 2
User rapidly able to identify existing
User Interfaces, Low/No code

Assess analytical
capabilities or composable opportunities
Vendor Vendor solution required
Deep learning, Predictive modelling, AI to solve the problem
composed services
3 4
Interfaces/APIs
Vendor Vendor Engage data Analytical solution democratised, less

Data pipelines, Data engineering
1 5 team engagement with specialists required
Cloud Cloud Collaboration with data team improves,

Data warehouse, Graph, SQL databases
1 2
Data team builds time is significantly reduced if bespoke
solution solution needed, drives innovation and
expose solution in user friendly manner
Extract, transform, load processes Vendor N
The user carries out the analytics

Data team carries
themselves through low/no code
out analytics
Composed
interface
services build
Data Source 1 Data Source 2 Data Source N from multiple
User with understanding of business
capabilities User receives
problem is closer to and helps shape the
output
output with developers
DIY Analytics – Distributed Analytics
Larger organisations that want to scale their data & analytics capabilities at pace should start to consider
distributed analytics strategies
BENEFITS OF DISTRIBUTED ANALYTICS

DRIVERS FOR DISTRIBUTED
ANALYTICS Agility: Decentralised teams can be more agile and operate independently,
reducing time to deploy and cutting down IT dependence
• Larger organisations that encounter more uncertainty and
changes in operations
• Organisations shifting towards becoming more distributed & Innovation: Domains are able to focus on their data products, bring new
decentralised enterprises data sources and solutions that otherwise wouldn’t have been thought of
• Organisations that want to scale quickly
Quality Control: Overarching governance principles encourage teams to
• Organisations that work with wide and frequently changing
deliver high quality data products in a common format
data sets
• Organisations that have legacy data platforms that are slowing
down innovation efforts Faster Product Delivery: Self-serve platforms provided by the data mesh
infrastructure handle complexities like data storage & management
• Organisations looking to move away from big centralised data
teams and monolithic data lakes or databases
• Companies that already have domain driven development Domain independence: Business functions have independence over
techniques prioritisation and data products that fit their needs
• Companies that have already started working with
micro services Cross functional collaboration: A data mesh encourages collaboration
between data experts on the business function side and developer teams
• Organisations with a culture that makes it difficult to scale
centralised services
Users close to development: The understanding of the operations and
business side of data is as close to the development as possible

DIY Analytics – The Data Mesh
Compared to composable D&A, a data mesh is a strategic approach to data management. Shift your
organisation beyond centralised data methods to a decentralised one by empowering data producers and
consumers to manage & access data without a central team
DOMAIN ORIENTED OWNERSHIP SELF SERVICE DATA FEDERATED COMPUTATIONAL

DATA AS A PRODUCT
& ARCHITECHTURE INFRASTRUCTURE AS A PLATFORM GOVERNANCE
• A domain is a group of people • This is where the value is drawn • A domain is a group of people • Data governance standards are
organised around a common from the data for the users and organised around a common defined centrally, but data
business function data products are created and business function governance is applied at the
• Here, we propose that the owned by the various domains • Here, we propose that the domain domain level in a decentralised
domain is responsible for the across your organisation is responsible for the management approach
management of data created by • They are self contained and each of data created by the business • Domains have autonomy over
the business function itself domain is responsible for the function itself data governance specific to their
• They are responsible for the security and management of their • They are responsible for the own products while adhering to a
composition, transformation and products' ensuring data is composition, transformation and set of global rules that ensure a
provision of data to end users, relevant, secure and up to date provision of data to end users, joined-up approach
eventually exposing the data as • They have a clear line of eventually exposing the data as • Autonomous data domain teams
data products ownership and can be consumed data products and centralised data governance
• The entire data lifecycle is owned by other data products or end • The entire data lifecycle is owned functions collaborate in order to
by a domain users to support business by a domain best meet the data needs of the
intelligence and AI organisation

Digital Ethics and AI
Digital Ethics can be embedded in the entire organisation from the strategic to the operational
level to ensure insights are trustworthy and technology aligns with company values
Responsible Innovation Strategy

Strategy Aligning values and vision through Digital Ethics guides
Digital Ethics responsible business strategy
Digital Ethics ensures that digital solutions are designed

and implemented in a responsible way.
It encourages organisations to align their processes and
Digital Ethics Governance
practices with relevant values and principles to safeguard
innovation and mitigate potential harm to stakeholders Governance Embedding Digital Ethics principles and governance
structures ensures successful implementation of
Digital Ethics considerations follow a solution throughout strategy and future-proofs ROI
its lifecycle to monitor the technology and avoid risks of
downstream consequences
It establishes accountability for how technologies are
used and designed across the entire organisation, creating Digital Ethics Culture and Values
processes to explain when it proves unsuccessful Digital ethics Instilling a Digital Ethics culture promotes inclusivity and
This generates responsible insights, allowing an culture effectively leverages diverse talent to create responsible
organisation to build trust and improve regulatory technologies
readiness to streamline implementation to
unlock growth
Responsible Insights: Smart, Scalable AI

Responsible
Centring Digital Ethics as the focal point of design and
insights implementation processes yields trusted insights

Responsible Innovation Strategy: A Digital Ethics lens uncovers the long-term implications
of AI for the organisation and its stakeholders
Questions leaders can ask when refreshing

strategy
The value of a digital-ethics Strategy
led strategy How can we future-proof the implementation

• The leading innovative organisations of tomorrow align of our strategy?
their actions to their values
While algorithms may be susceptible to risks with Digital Ethics
• A digital ethics-forward strategy can leverage your ESG implications, when combined with appropriate human oversight
goals to drive more resilient strategy development they can provide many benefits to society and wider-society
Governance
• A strategy centered around digital ethics considerations
starts with determining what matters to stakeholders;
customers, employees, regulators, and the general How do our values guide technology selection
public to guide responsible business and adoption?
• This builds trust with consumers and business A responsible innovation strategy should cascade through use
partners that fosters collaboration for more productive case selection that aligns with strategic priorities which flow
Digital ethics
and efficient relationships that will ultimately benefit from organisational values and differentiated based on
culture
compatibility
everyone involved
• Simply meeting the regulatory standards is not enough –
proactively addressing digital ethics considerations How can we build an inclusive “digital” brand?
grows brand trust and prevents reputational damage
Becoming a digitally responsible brand starts with ensuring
from the use of data in business
digital ethics is a part of the DNA of the organisation and
Responsible
effectively communicating internally and externally how the
insights
organisation is using digital ethics to be a more responsible
business

Introducing Digital Ethics governance that is principles-based is an effective first step towards
responsible use of AI
Governing technical systems with digital ethics Governance principles
Creating Digital Ethics governance structures based on
principles is an important step in enabling an organisation Human- Value-sensitive design creates internal and external checks to help
to effectively deploy technology to produce responsible Strategy
Centered & Fair ensure equitable application of technology across all participants
insights. Governance policies should consider three key
aspects when operationalising governance principles:
Robust and AI systems have the ability to learn from humans and other systems
Accountability and compliance
Reliable and produce consistent and reliable outputs
Digital Ethics governance identifies ownership
and accountability chains across the
organisation, instituting Digital Ethics advisory All participants are able to understand how their data is being used and
Transparent & Governance
capabilities in existing governance processes how technology systems make decisions; algorithms, attributes, and
allowing escalation to board-level as required. Explainable correlations are open to inspection
This ensures organisations can meet regulatory
obligations and manage reputational risk
Consumer privacy is respected, and customer data is not used beyond
Privacy its intended and stated use; consumers are able to opt in/out of sharing
Aligning actions with values their data
Digital Ethics governance increases robustness
and confidence in internal processes to drive Digital ethics
responsible business practices. Governance Responsible & Policies are in place to protect human agency and determine who is culture
allows organisations to align their actions with Accountable held responsible for the output of AI system decisions
their values
Mitigating sociotechnical impact and AI systems can be protected from risks (including Cyber) that may cause
building trust Safe & Secure
physical and/or digital harm
Unlocking the opportunity of technology requires
also addressing the risks posed to internal and Responsible
external stakeholders, including the public. Inclusive & Systems are designed to be inclusive and ensure individual, societal and insights
Proactive governance builds trust and ensures Sustainable environmental wellbeing
responsible technology use

Embedding digital ethics principles in governance structures is critical to the reputation,
brand trust, risk mitigation and effective monetisation of AI capabilities
How the digital ethics governance is implemented depends on the organisation, the technology in use, the context and its IDO maturity. The framework can
help address the organisation’s key pain points around ensuring insights are trustworthy
Strategy
How can industry apply the framework? What are the benefits?
How to Exceed Compliance Competitive Edge

With increasingly digitalised economies, regulators are pushing The use of AI is highly restricted in many industries. Unpacking
more and more towards safe and secure AI technologies and are black-box AI models and putting in place proper frameworks
imposing guidelines and regulations on trustworthy AI. Digital dealing with AI systems enables companies to use artificial Governance
Ethics governance that proactively addresses these issues ensure intelligence at the core of their business and go new ways where it
regulatory readiness, safeguarding your investment was not possible before
How to Build Reputation/Trust

Would you trust an AI with your health or life savings? Many
Reduce Risk
companies have developed good AI systems. Users however, Making trustworthy AI a core competency in your business helps
struggle with trusting their results. Trust in AI, is more than model to understand and address the vulnerabilities arising from Digital ethics
accuracy. It is also defined by how the AI came to its conclusion, adversarial model behaviour as well as arising risk of human culture
how it influences decisions and who takes the consequences misbehaviour when interacting with AI systems
when something goes wrong
How to Generate Responsible Insights Improve Revenue

Digital Ethics considerations such as transparency and
Generating responsible insights open the doors to numerous
explainability when formalised through governance processes Responsible
business opportunities in the artificial intelligence space, which
highlight aspects such as understanding how a model comes to its insights
market size was valued at $ 39.9 B in 2019 and is expected to
conclusions and where it has its weaknesses that encourage fine
grow by 42.2% annually from 2020 to 2027
tuning of models and boost model performance

Digital Ethics Culture and Values: Instill a reflective decision-making process
A culture of reflection
A Digital Ethics culture encourages iterative reflection throughout the analytics and AI
delivery lifecycle to ensure risks are anticipated, identified, escalated Strategy
appropriately and mitigated
DOING THE
A culture of transparency
RIGHT THING. Stakeholders, including customers, often have high levels of trust in organisations that
TOGETHER. are open and transparent about their use of digital solutions. Ensuring external
stakeholders are aware of how the organisation uses technology and what impact they
can expect is critical to building trust
Any new decision, model, product Governance
The impact of new solutions should be understood and communicated internally as
or service comes with risk as well well, to build awareness on how the organisation aligns its actions with its values.
as opportunity Responsible teams are adaptable, resilient and proactive
With advances in AI, it is
important not to underestimate A culture of accountability
the need for human and AI When having clear accountability chains is a priority and embedded in the culture of an
balance to mitigate harm and organisation, customers, employees and others will be able to seek redress in the case Digital ethics
reap the benefits of innovation they suffer harm from a digital solution. Demonstrating accountability builds a trusted culture
relationship with all stakeholders
Ultimately, an organisation needs
to be comfortable with the A culture of inclusivity and diversity
societal impact of their Creating diverse teams (e.g. different ethnicities, genders, disabilities, ideologies, and
technology and how that belief systems) at all levels of the organisation drives resilience and provides a broader
judgment call is made base of perspectives to identify and mitigate problems and risks such as bias in data
and models Responsible
insights
…A DIGITAL ETHICS CULTURE

Responsible Insights: How operationalizing Digital Ethics delivers Smart, Responsible, Scalable AI
Smart scalable AI Smarter

Digital Ethics considerations promote Strategy
AI that is more trusted, resilient,
Smarter, more responsible AI disruptions that scale to creative and less data-hungry
your long-term vision
Highlighting the ethically risky elements
Machine learning and AI have seen significant expansion in recent of AI applications enables discovery of
years, with advanced analytics being embedded into a broad range of technical solutions
processes and applications; from CRM systems driving enhanced
customer experience to back-office systems to reduce fraud and drive Governance
efficiency.
As AI becomes more central to all aspects of society and business, there
is also a growing urgency to protect ethical use and privacy when
Scalable
using AI. Ethical guidelines are important to establish as bias and Embedding Digital Ethics in the
discrimination in decision making are amplified when automated with AI. production process improves AI
Current AI benefits can be difficult to achieve in the face of new and robustness, accessibility, productivity, Digital ethics
evolving issues that accompany deployment of a production ML deployment and operations ensuring culture
application. Machine learning applications encounter snags that other risks are methodically addressed as the
software projects do not. Resulting in AI projects incurring delays organisation matures
increasing costs and postponing benefits.
Digital Ethics can facilitate safe and successful implementation
acting in tandem with the development process; utilizing Digital Ethics Responsible
considerations to either signal issues to be addressed or to provide a insights
Digital Ethics dimension to reinforce a solution

Responsible Insights: Digital Ethics considerations highlight the ethically risky elements of AI applications,
enabling discovery of appropriate and effective technical solutions
Smarter AI – A Case Study: Overcoming Ethical Data Limitations with Synthetic Data
To overcome data availability concerns organisations looking to leverage data-intensive AI can opt for cutting edge solutions like synthetic data to meet innovation goals and mitigate
stakeholder harm. Data security and privacy are key concerns for organisations who desire a high degree of stakeholder trust by decreasing the risk of exposing sensitive data to new
vulnerabilities Strategy
Three main trends to meet the challenges of provisioning test data:
Trend 3: Synthetic data Coverage

Trend 1: Data Subsetting
Enhancing test coverage by mirroring production data and filling in
Using a subset of production data in “gaps” with richer datasets the first time
Combining the benefits
non-production environments E.g. A Large Bank may generate large volumes of data to cover many test
✓ Distribution and testing faster* Generating synthetic data cases without compromising its current member’s identities
Governance
✓ Most stable trend combines the benefits of both
trends through synthetic Security
x Not representative of entire dataset
representation of the full Avoids exposing sensitive data in testing, such as Personally
x May not retain referential integrity* Identifiable Information (PII)
production database without
x Exposes production data to testers* E.g. A Hospital Network can use synthetic data to protect patient privacy
compromising sensitive data.
x Requires data masking of PII* in compliance with HIPAA while developing online health portals
x Potential misuse/breach
This allows organizations to Agility
ensure key Trustworthy AI Automatically creates new data, and effective, efficient distribution Digital ethics
governance considerations are to testers, without relying on operations assistance culture
Trend 2: Data Virtualisation
met: E.g. A Large-scale Enterprise can scale this solution to create large
Creating virtual copies of production data in non-
• Monitoring and Oversight volumes of data for accurate performance, load and regression testing
production environments
• Data Minimisation before going to market
✓ Representative of entire dataset
• Maintaining transparency Maturity
✓ Increase in vendors offering data virtualisation
• Scaling governance Tools now support complex environments and quick data access to
x Less mature than other trends keep pace with more rigorous development practices Responsible
E.g. An organisation using DevOps and Continuous Integration can use insights
x Misunderstanding of data/coverage this to compliment their mature processes
* Applies to trend 1 and 2

Responsible Insights: Incorporating Digital Ethics into the production process builds trust by
embedding a human-centric focus throughout the products lifecycle to systematically mitigate harm
Scalable AI – A Case Study: Revolutionising ML Development with Digital Ethics* MLOps
MLOps enables organisations building and deploying AI applications to decrease production delays, ballooning costs and missed benefits
Digital Ethics seamlessly embeds itself in this process; Safeguarding project value by creating AI that is resilient, reproducible and robust, ensuring risks are proactively addressed
in an agile, timely manner for ease of implementation and trusted results Strategy
Model Packaging and Deployment simplifies deployment of models built using a variety of languages at scale on any cloud or on-premise. Successful
user acceptance testing in preproduction triggers model deployment and integration with downstream applications.Governance tools are in place
and appropriate communications to relevant stakeholders are cascaded
Model Monitoring and Management enables continuous tracking of various performance measures against appropriate thresholds and
benchmarks identifying retraining needs. Business ROI and other business metrics are also tracked to demonstratebenefits realisation
Governance
Model Governance includes features such as model access control, an audit trail to provide transparency on the model’s functioning and
any other regulatory and compliance needs for model usage
Model Security ensures the protection of models from being exposed to cyberattacks or inappropriately accessed by unauthorised users
Digital ethics
Model Discovery provides model registries or catalogs for models produced within the tool ecosystem as well as a searchable model culture
marketplace that provides a way to locate consumable models, both internally developed as well as third-party models which are vetted
against the Trustworthy AI framework
Data Preparation ensures data governance practices are enacted to collect, label, cleanse and process appropriate data and that these
measures are repeatable as the model is deployed
Responsible
ML Pipeline Development ensures use cases are selected and prioritised in line with the organisation’s strategies and values. It allows users to insights
define repeatable and reusable steps for model development
*green text highlights Digital Ethics measures

Start small, scale fast
As companies experience the limitations of big data as a critical enabler of analytics and AI, new approaches
known as “small” and “wide” data are emerging
Small and wide data enable organisations to drive value faster
With technology rapidly evolving enabling businesses to store more data and scale fast,
one might think the solution is even bigger data. However, experts believe the opposite Wide Data
– companies have struggled to gain meaningful insights at speed where big data has
X analytics of video, text, image, audio,
clouded decision making and limited building bigger picture ideas and responding to fast
tabular, json, temperature, …
moving market shifts.
Disruptions such as the COVID-19 pandemic can cause additional problems because
historical data captured in big datasets become obsolete more quickly.
Small and wide data provides organisations the opportunity to gather more specific,
relevant and distinct insights from a multitude of data components where big data has
struggled to do so.
• Wide data focusses on tying together disparate data sources such as video, text,
imagery and many other data types – this helps organisations generate more
meaningful and contextualised analysis. Predictive maintenance of engines is a good
example for wide data – a variety of data types are brought together to generate Small Data
insights that forecasts engine failure Time series analysis, snapshot learning,
federated learning, adaptive, …
• Small data is what it says on the tin – using smaller datasets that enable organisations
to drive value quickly and work with agility to solve problems – very useful if datasets
need to be understood quickly with an iterative approach – many AI techniques are now
being developed to cope with small data sets

Small and Wide Data allow for more robust analytics and AI, at pace
Context Innovation 360 view

Predict passengers on train by Reduce time and effort needed to Derive better decisions by building the
combining sensor, timings, schedule, deploy new analytical tools and data full picture from multiple data types and
train capacity data. Add more context hungry barriers for ML models. sources. Improve situational awareness
for better predictions! Empower decision intelligence
Robustness Agility Pace

Forecast demand with frequently Build, deploy, learn and fail fast. Build Lower time barrier to entry for advanced
changing customer sentiments iteratively and adapt easily – at analytics without requiring complex
incorporating wider and smaller data reduced costs! build of data storage architectures
into less data hungry models – remove
bias of historical observations

Big data shouldn’t be discarded, but the benefits of small and wide data are clear
Benefit Small data Wide data

Usually structured, easy to manage
Easy to use for testing and hypothesising
Use to show value at pace and low cost
Add contextual value to AI and analytics
Use in X Analytics*
Extract value from all enterprise data
Helps solve the age-old data silo problem
Proactively respond to changes in market conditions with agility
Critical to predictive analytics
Generate mores specific and creative insights with AI
Stay more personal
Get real time insights
* X Analytics is the ability to run any type of analytics on all of an organisation’s structured and unstructured data, no matter where or in what format the data resides

Big Data is all about finding correlations by analysing the past. Small data gives organisations the
opportunity to find the causation, the reason why
Small data is often best used by asking Techniques are being developed to build analytics and AI capability
structured crunchy questions from small datasets
• Organisations today are often collecting large amounts of

data and find it difficult to leverage value or know where Gen-AI systems are capable of outputting novel text or creative
to start Generative media content using existing text, audio video or imagery. These
• A structured approach involving small data is often the best AI models can be trained on small data – e.g., a music album – to
way to start generate create content specific to the use case required
• Organisations should begin by asking the right questions that

can be grouped into logical themes which form analytical use
cases, from MI to predictive models – such as:
Synthetic Use small data and extend the data by synthesising computer
“How can we use external and internal data to cross reference Data generated data that preserves the original data’s statistical features
against target demographics to best locate a shortlist of
locations to cut search time by 50%?”
• These often need specific datasets that one may find difficult
to obtain from big data sources. Time series Time series forecasting can be applied to smaller datasets so long as
the data displays seasonality and trend that can be learnt by ML
analysis
• These can be tested with small data using techniques such algorithms
as generative AI, few shot learning, synthetic data and
federated learning – these help build analytics and AI POCs
quickly without requiring full blown big data architectures
Few shot learning is a type of ML method where training datasets
• These can then be prioritised and delivered into production Few shot contain limited information. It aims to build accurate ML models with
at scale once your organisation has understood value using learning less training data – applications include computer vision, natural
small datasets first language processing, object recognition, translation and more

Organisations can benefit most from small and wide data by leveraging the right process and
governance framework
• Sometimes, breaking down both problems and solutions into bite
size chunks lets you derive more valuable, specific and distinct insights Step 6
from your data at pace – rather than clouding thoughts and creativity by
Deploy and scale
relying entirely on the processing of big data and complex data hungry
in production
deep learning models
Step 5 • Upon proving value,
• Collecting large volumes of data can be a challenge for organisations –
organisations can choose
what if you could start narrow and scale upon proven success? Prioritise and iterate to invest the time and
• It is important to note from the below framework – small and wide data with agile money into deploying
does not fall in the face of robust data governance! Step 4 • Prioritise POCs and use production grade
cases according to architecture, building
Test with small and wide business value and necessary pipelines and
data complexity completing full data set
• Build proof of concepts integrations
Step 3 • Iterate on a minimum
at pace and low cost viable product with agile
Analytic Use Cases using small and wide methods, building on
• Build the analytical use datasets capability until value is
Step 2 cases that can then be • Start small and wide, proven and met
executed as proof of learn fast (fail fast!)
Ask the Crunchy Questions concepts
• The crunchy questions • Park the questions which
• This can be the build of cannot be answered or
Step 1 break the problem into an AI model to answer a
small chunks are incorrect hypotheses
Start with the problem crunchy question
• Questions should be
• Start with defining the specific, narrow and
problem and the answerable quantifiably
business ideas

Analytics on the edge
Computing environments residing closer to assets in the physical world and outside IT’s typical purview
Increasingly, data and analytics and the technologies supporting them reside outside traditional data centres and cloud environments. This is where edge
comes in: Deloitte defines edge computing as distributing functionality, such as computer processing and data storage, so that it’s located closer to— or
even embedded in—devices that produce data.
Edge continuum
Cloud/data Endpoint
center devices
Traditional Real-time operational Analytics-enabled AI-enabled

analytics and BI intelligence autonomous “smart products”
As this trend continues, there is both a requirement and a huge opportunity for organisations to enable greater flexibility in how and where data
management and analytics are carried out. These changes will significantly impact D&A leaders and their teams, requiring new capabilities and skills while
also opening up new opportunities to deliver value.

Edge computing unlocks the opportunity for multiple benefits to organisations
Speed and Lower Latency Reliability and Resilience
By distributing data and analytics capabilities to edge environments, Assets driven by edge environments can continue to operate on
the need to move data from endpoints to the cloud and back again is their own even when communication channels are slow,
eliminated. intermittently available, or temporarily down.
Scalability Security and Privacy
With an established edge computing environment, the network can Edge computing keeps data close to the edge instead of in
be extended by increasing the number of devices, data centres, centralised servers, holding limited amounts of data and
and processors, without affecting other parts of the network. incomplete data sets that are less compromising if hacked.
Cost Effectiveness
As data does not need to travel back to centralised servers for

devices to function, edge computing provides savings in
operational costs and reduces storage needs.

Computing at the edge powers new applications
Like cloud computing, edge computing can galvanise new applications and business models. Processing and analysing data at the edge reduces the
distance data must travel, producing insights faster and sidestepping the privacy concerns and costs associated with transmitting and storing data in the
cloud. Significant long-term cost savings can be realised leveraging edge-based processing instead of the cloud.
Edge computing use-case category Representative use cases
• Room occupancy sensors or cameras in smart buildings

• Signals processing (e.g., detecting problems in infrastructure, such as
Insights: Generate insights that cloud-based analytics cannot deliver because
oil pipelines)
of latency, lack of connectivity in remote areas, privacy concerns, or other
• Condition-based prescriptive and/or predictive maintenance
limitations of cloud computing
• Thermal/temperature monitoring
• Medical imaging
• Augmented reality (AR) for object status and maintenance
• AR for in-store shopping
Experiences: Delivering highly responsive and immersive digital interactions
• Vibrating warning (e.g., on a steering wheel for driver assistance)
or experiences
• Virtual reality (VR) and AR training
• Multiplayer immersive gaming
• Remote-controlled drones for aerial inspections
Remote control: Enabling remote personnel to coordinate actions and • Remote fleet management for vehicle operations
responses
• Industrial automation (e.g., manufacturing or distribution centre

Automation: Enabling devices to respond to and act on information control systems)
generated from their surroundings • Autonomous cars
• Home automation (e.g., smart speakers)

Understanding edge computing architecture
Edge computing architectures can involve processing data at multiple levels. Processing at different points has both technical and use case implications.
Where the data is processed Description (typical latency) Real world example
On-device Data is processed on the device that generates it, such as a Medtronic's Guardian Sensor 3 leverages on-device processing
smart sensor, smartphone, or connected vehicle. to update users on their blood glucose levels every five
minutes.
Processing at the “on-prem” edge Data generated by a device is transmitted to and processed on Cree uses edge servers from Microsoft optimised for AI
(or local area network edge) another nearby device, such as a network gateway or server, inference to support real-time image analysis for quality control
that provides compute for devices within a small geographic in its manufacturing facilities.
area, like a warehouse or industrial facility.
Processing at the “near” edge Data is transmitted to and processed at a network gateway, AT&T and Dell are working together on open source technology
(or wide area network edge) such as a cell tower or switching station, for a wireless network that can automatically provision edge computing resources at
that covers a wide geographic area, like a whole city. cellular network sites to support various edge applications.
Processing at the “far” edge Data is transmitted to a physically proximate data centre AWS recently announced Local Zones, an initiative to establish
(a local data centre) specifically established to provide low-latency processing small data centres near metropolitan areas where enterprises
for devices and applications in its vicinity. can deploy low-latency applications closer to their customers.
The first Local Zone went live in Los Angeles late last year.

Secure collaboration
Powerful data-sharing and privacy-preserving technologies usher in a new era of data monetisation
NEW TECHNOLOGIES FUEL DATA SHARING TREND
You can now buy and sell valuable information assets in highly efficient, cloud-based marketplaces. Combining this data with a new array of privacy-preserving
technologies provides the best of all potential worlds: sharing data while preserving security and privacy.
Stores of sensitive data lying fallow in servers around the globe due to privacy or regulatory concerns are starting to generate value across enterprises in the form of
new business models and opportunities. During the next 18 to 24 months, we expect to see more organisations explore opportunities to create seamless, secure data-
sharing capabilities that can help them monetise their own information assets and accomplish business goals using other people’s data.
Though currently in an early stage, this data-sharing trend is picking up steam. In a recent survey, Forrester Research found that more than 70% of global data and
analytics decision-makers are expanding their ability to use external data, and another 17% plan to do so within the next 12 months.
More than 70% of global data and

analytics decision-makers are expanding
their ability to use external data

Shared data leads to greater value and therefore higher business performance
The rapid growth in data sharing can be attributed to the simple fact that data gains value when it is shared.
In fact, Gartner™ predicts that by 2023, organisations that promote data-sharing will outperform their peers in most business metrics
DATA-SHARING EXAMPLES
1. Using aggregated data 2. Increasing efficiency 3. Broadening your 4. Securing intellectual 5. Encrypting data in
to securely achieve and lowering costs research property motion
common goals collaboration
Organisations can work with Across enterprises, data Sharing basic foundational Super-sensitive data such as In the arenas of high-
“frenemies” within a market vendors no longer have to or early-stage findings can AI training data that may be frequency trading, robotic
sector to achieve common provision hardware, accelerate critical research stored in public clouds can surgery, and smart factory
goals such as developing maintain databases, and initiatives without be better protected. manufacturing, confidential
deeper customer insights or build application compromising a hard-won data flows rapidly across
detecting fraud patterns programming interfaces competitive advantage. multiple entities.
across an entire sector. (APIs). Customers can push Homomorphic encryption
a button to access allows users to access
anonymised, curated data critical data quickly without
feeds. Within the enterprise, encryption keys.
encrypted data makes
artificial intelligence (AI) and
machine learning (ML)
exercises safer, and
compliance audits easier.

Emerging technologies that help preserve privacy while sharing data
In order to truly capture value from data sharing, institutions need to be able to use the data available to them, both within their institutions and outside of them.
Below are five key techniques for preserving data privacy that make it possible for organisations to reap the benefits of data sharing without sacrificing privacy.
Federated Analysis
Where parties share the insights from the analysis of their data without sharing the data itself
Zero-knowledge Proofs
Where users can prove their knowledge of a value without revealing the value itself
Differential Privacy
Where noise is added to a dataset so that it is impossible to reverse-engineer the individual inputs
Homomorphic Encryption
Where data is encrypted before sharing, such that it can be analysed but not decoded into the original information
Secure Multiparty Computation

Where data analysis is spread across multiple parties such that no individual party can see the complete set of inputs

Emerging technologies that enhance privacy and confidentiality
Technique Definition When it’s Useful Example
Without differential privacy: With differential privacy:

A third party knows the spend of several others One of the inputs is removed and replaced with a
and the group average. random figure.
Differential privacy allows 7K

Where noise is added to a
aggregate information on a
Differential dataset so that it is impossible 4K 7K 6K 5.5K 6K 6K 4K 7K 6K 5.5K 6K 6K
dataset to be publicly shared
Privacy to reverse engineer the
whilst also protecting individual
individual inputs.
information within the dataset
The third party can The shared “group average’ is noisy, making it
find out John’s spend. impossible to reverse-engineer John’s spend.
With federated analysis:

Insurer A
The person named “John McScammer” Shared fraud detection engine:
Sometimes data needed is has committed fraud in the past
scattered across multiple sources • The person named “John McScammer”
Where parties share the insights
Federated and combining sources may not Insurer B has committed fraud in the past
from the analysis of their data
Analysis be possible due to different Owners of green cars are more likely
without sharing the data itself. to commit registration fraud • Owners of green cars are more likely to
jurisdictions or privacy
restrictions. commit registration fraud
Insurer C
Drivers living in the 12345 postal code • Drivers living in the 12345 postal code
are more likely to commit claims fraud are more likely to commit claims fraud

Without homomorphic encryption: With homomorphic encryption:
John places his health records in a box, ships them to the John’s health records are homomorphically encrypted prior
company, which analyses them to produce a report and to sharing, making it difficult for anyone but him to see the
Sometimes a company would like ships it back to John. data or the result of any subsequent analysis.
to engage a third party for data Data could be malicioulsy Data is secure during
Where data is encrypted before analysis using complementary accessed in transportation. transportation.
Homomorphic sharing, such that it can be data or proprietary analytics the
Encryption analysed but not decoded into company doesn’t have. However, Insurance Co. Insurance Co.
the original information. the data steward or owner may
lack permission to transfer the
data or have privacy concerns.
Data cloud be maliciously accessed by the company

itself or an external bad actor gains access to the office. The company conducts its analysis without being
able to see the underlying data at any point.
Without zero-knowledge proofs: With zero-knowledge proofs:

“Does your income meet
When customers would rather my requirements?”
Attributes: “Does your income meet
not reveal more than is absolutely my requirements?”
Attributes:
Zero Where users can prove their Age
necessary to complete a Age
Knowledge knowledge of a value without
transaction with the worry that “Yes, my income is £80K” Income “Yes”
Proofs revealing the value itself.
the information will be used ZKP system Income
against them. Gender
(mathematical process that Gender
can be independently verified)

Without SMC With SMC
Secure multiparty computation

allows institutions to jointly
analyse data without one Hedge fund Data provider Hedge fund Data provider
Secure Where data analysis is spread
institution being able to access Confidential Private Confidential Private
across multiple parties such that
Multiparty the complete dataset. This allows models data models data
no individual party can see the
Computation multiple institutions with sensitive
complete set of inputs.
information to work together to
create value without risking Sensitive data must Sensitive data cannot
confidential information. “The use of this data would be shared directly “The use of this data would be accessed by
increase returns by 2.4%” with counterparty. increase returns by 2.4%” counterparty.

Making the Right Impact

War for talent
Creating an ecosystem of talent which will facilitate distributed enterprise
DIGITAL READINESS ENGAGEMENT AND RETENTION
Organisations need to adapt to a rapidly changing era of continued digitisation and Developing and growing talent is important to increasing the breadth and depth of AI
enterprise distribution. Transition to a digital centric operational model requires a capabilities. Defining specialist learning and developing pathways for your talent, in
certain set of skills to implement the processes. addition to initiatives for knowledge sharing, both within your capabilities, as well as
looking at opportunities externally to knowledge share keeps skills and ideas fresh
Organisations need to rethink skill management strategy at times of tech innovation
and relevant.
and how to address perishable skills (with a short life span). Focus on developing
proactive training plans, instead of reactive chaotic activities plugging gaps. ‘Data literacy’ is also important for business users, training pathways can open up
opportunities for both technical and business teams to develop skills progressively
Industries are encouraged to collaborate with educational organisations to develop
and continuously.
programs for upskilling existing employees. This will allow internal career growth and
contribute to talent expansion and retention within an organisation.
THE BATTLE FOR DATA SCIENTISTS DEVELOPING A PIPELINE FOR ANALYTICS TALENT
The battle tho secure and retain data science talent in real; there’s simply not enough To attract and retain the best talent and continue to keep their skills leading edge,
data literate and digital savvy talent to fill the jobs. Knowing what motivates data organisations need to invest in building capabilities long before employment.
scientists to change jobs and promoting the correct growth culture can provide Organisations are teaming with leading universities and professional associations
valuable insight on how to nurture this talent in your team. within their ecosystems to build the next set of AI talent:
Building insight driven teams who work closely together and are encouraged to pick • Designing curriculums and teaching courses
up skills from other members facilitates flexibility; when one member is unavailable,
• Hosting conferences and working on joint training
others can step in. This helps create a team that is more likely to remain with the
organisation – when everyone is learning new skills from their team and advancing • Recruiting into our practices and referring to clients
their careers, team members have more reasons to stick around.
• Investing in start-up side businesses to encourage employee innovation

War for talent
Recruiting the right talent to drive innovation
WHO ARE WE FIGHTING FOR?
In the era of the distributed enterprise* the demand for data scientists, data
architects and data engineers soars, while supply of complex digital skillsets remain
short. Specialised roles which cannot be built internally are difficult to acquire and
require new approaches in talent sourcing. Based on the outcomes of global IDO
survey** 55% of respondents expressed the need for AI skills that they cannot
source – over 10% increase on last year. This was also reflected for data science skills,
as 60% respondents do not use AI in any of their business functions predominantly
due to a lack of priority and skillsets.
While there is a demand for more niche AI focused roles such as ML engineers,
ultimately data scientists remain in high demand due to the diversity of their
expertise. The data scientist definition speaks to a rare blend of statistical
sophistication, data management skills and business acumen. One of their core skills
is the ability to extract meaning from raw data and use that knowledge to drive
business transformation with insight-led decisions, innovation and operational
efficiency. Today, both the private and public sectors need a new breed of data
scientists who can handle sophisticated data analysis and navigate big data freely,
but who also have fluent communication skills, business acumen and political nous.
Organisations often limit themselves by trying to standardise data scientist job
descriptions, despite the fact that the role covers a diverse set of skills and
candidates can originate from all walks of life. Data scientists can take on multiple AI
related roles within an organisation, unlike ML engineers who primarily focus on
data modelling.
© 2022 Deloitte LLP. All rights reserved. IDO Playbook 2022 76 76

*See topic “Distributed Enterprise” **Global IDO Survey 263 total respondents across 17 countries IDO Playbook 2022 76
War for talent
Focusing on the right balance of business and technical skills
CORE CAPABILITIES
Being able to blend business and technical skills is critical for the success of an AI capability. Very rarely, a blend of skills may be present in a very highly skilled individual;
however, it is this blend of blue (business) and red (technical) skills within a capability which drives purple talent. Consider creating an Insight Driven Team for every AI
project you embark on.
BUSINESS AND COMMUNICATION TECHNICAL AND ANALYTICAL SKILLS
• Agile, Hybrid, Project • Innovation Management • Advanced Visualisation • External Data Provisioning
Management
• Knowledge Management • Machine Learning Engineering • Infrastructure Management
• Budgeting and Funding
• Market Analysis • Configuration and Release • Load Testing
• Business Analysis Management
• New Service Design • Natural Language Processing
• Business Change • Customer Segmentation
• Partnerships and SLA • Neural Networks
• Business Metric Definition Management • Data Mining
• Regression Modelling
• Capacity Management and • Process Design • Data Protection and
• Security Advisory
Scheduling Anonymization
• Stakeholder Management
• Server Administration
• Community Management • Data Quality Enhancement
• Storytelling
• Solution Architecture
• Demand Prioritisation • Enterprise Architecture
• Training
• Unstructured Data Processing
• Functional Testing and UAT • ETL Development
• User Experience Design
• Web Analytics
• Incident Management • Cyber Security
• Vendor Assessment
• AI Deployment
• Data Architecture

War for talent
Adopting a human-centric hybrid model to attract and retain talent
ATTRACT, RETAIN AND UPSKILL TALENT SOURCING
Unearthing existing talent within an organisation can be a benefit for all involved; achieve Organisations should recognise that success depends on more than technology talent.
individual career ambitions within AI, reduce the organisational resource management For example, data scientists often struggle when they aren’t clear on the business
costs of recruiting highly skilled talent, and retain an employee who pursues AI ambitions. problem they’re supposed to solve. The result can be AI projects that goes nowhere, and
disillusioned data scientists who deflect to competitors. Subject matter experts who can
Getting the right motivational factors in play is fundamental to successful long-term
‘speak data’ to data scientists while ‘speaking business’ to executives are invaluable.
retention; high degrees of autonomy, opportunities to develop as subject matter experts,
and above all a clear purpose to the work at hand.
SKILL CRUNCH & CROSS-SKILL
Facilitate acquisition of new skills and competencies which can be implemented in

multiple job functions and will allow team flexibility. This will increase overall productivity
and talent utilisation.
HUMAN-CENTRIC HYBRID MODEL UPSKILL
Physical working spaces such as a cockpit of screens or Bring Your Own Device are Elevate your employees' skills beyond adequate for the role which will allow wider
ranked highly by Next Generation talent. Agile is becoming a norm post-pandemic break utilisation of talent within the organisation. This can be done through interactive training
out. Employees in data and IT space are seeking better life-work balance, making a hybrid material, professional and industry related certification, coaching, webinars and
working culture one of the key drivers for talent acquisition and retention. Similarly, workshops and digital tool/platform learning.
organisational culture and innovation, such as Google’s famous encouragement of
Upskill management to address employee wellbeing demand and reduce talent attrition.
continuous innovation through empowering employees to improve processes they don’t
like, attract talent who enjoy thinking outside of the box.

War for talent
How to maintain and manage your talent pipeline
KNOWLEDGE MANAGEMENT
Managing a knowledge base is very important as AI teams grow and work

with an increasing breadth of data and deployed analytical models.
Using centralised tools to manage knowledge helps retain and share
knowledge amongst your analytics team(s) and more broadly across the
organisation. It is vital that organisations can distribute and scale these
tools enterprise wide.
Enabling employees, stakeholders and ecosystem partners to collaborate
better and work more efficiently through powerful, user-driven enterprise
tools is a core component in building insight capabilities. This often
involves re-imagining enterprise user interfaces and engagement models
to drive adoption, productivity, and stakeholder sentiment.
Some quick-win examples include:
• Enterprise social networking platforms
• Cloud-based collaboration platform
• Algorithm repositories
• Visualisation hubs
• Crowdsourcing communities
The ongoing management of the repository should also be considered to
ensure relevant and up-to-date material which engages the audience.
© 2022 Deloitte LLP. All rights reserved. IDO Playbook 2022 79 79

Adopting IDO Culture
Evolving the data culture of your organisation fuses automated, analytics driven insights with human
intelligence for an exponential impact
“Adopting insight driven culture across an organisation requires your business vision and data strategy to be built in parallel as one cohesive whole, continuously shifting
the paradigm as new data and technology trends emerge. This includes changing beliefs, values and behaviours – the 3 roots of culture” Deloitte Italy – IDO Lead
LET'S CHANGE OUR MINDSET
“
The importance of an insight driven culture is a key piece of the puzzle to success
that is often forgotten about.
Strategy incompatible
“
with an organisation’s culture will fail
Successful implementation impacts our behaviours, beliefs and values where every
person is willing to make decisions, change processes and adapt behaviours based
on insights rather than intuition or tradition.
Some of the blockers that can prevent organisations from adopting IDO Culture and
must be addressed include:
`
➢ Legacy technology including a lack of data governance
➢ Fear or lack of change
➢ Lack of trust or access to data
➢ Generational ways of working based on experience alone
Strategy outlines Culture is how that
the work needed work gets done
Adopting a data centric culture can be done through good communication of
benefits and a compelling picture of the future, advocates who can powerfully
articulate the need for change, active engagement of key figures to create a ‘pull’
from leadership to enable the change and mobilising the right team and skills to
design and support any solution with excellence.

Winning organisations configure an insight driven culture by strategically tailoring a well-defined cultural
maturity state that is deployed into action through small, measurable changes
3 ACTION CHANGE
7. Target priority areas of your organisation to activate

2 GATHER INSIGHT change; perform organisational analysis to identify
influencers who can drive viral change
8. Define and prioritise critical cultural behaviours
1 ALIGN TO STRATEGY 9. Use design thinking to tell the culture story at the local level
`
3. Assess current state of culture and data literacy 10. Conduct a hackathon to brainstorm interventions to activate
4. Enrich findings with focus groups, interviews your desired results
and crowd-sourcing 11. Establish minimum viable changes and build action plans
1. Conduct leader interviews: understand
mission, vision, value, strategy alignment 5. Conduct gap analysis between current culture 12. Execute short-term culture ‘sprints’, linking targeted
and gather input for future state culture state and desired future state behavioural interventions to value events
vision
6. Identify organisational levers to push/pull 13. Deploy culture enablement coaches to work directly with
2. Culture lab: define a compelling cultural in order to enhance or minimise cultural team
aspiration aligned to the organisation’s attributes using a culture maturity model
mission, vision, values, and strategy 14. Measure progress, iterate to arrive at the right approach
15. Establish accountability, governance and PMO processes

The most important part of a successful IDO cultural configuration is to uncover your
organisation’s culture levers and tailor Minimum Viable Changes (MVCs) to shift the dial
Identify ‘as-is’ performance across an organisation’s cultural lever and …Leverage enablers and blockers to define quality Minimum Viable
prioritise according to strategy… Changes for each lever
Agility, Constant
Iterative Disruption 1. Assess and select a commercial collaboration tool
2. Define collaboration tool metrics (e.g., monthly, weekly, daily active
Digital users) and establish tracking mechanism. Begin tracking progress
Decision
Velocity vs
Criteria 3. Inventory decision rights of managers to determine rights that must
Legacy
stay with managers and those that can be shared with staff.
Make decisions on what roles can have which rights.
Team Example 4. Revise selected decision rights for X group or X level
structures Culture Levers
Collaboration `
5. Create collaboration zones in current office space
6. With selected customer-facing team, change from customer ‘show
and tell’ to client-company co-development process. Design new co-
Customer Democratised development process. Pilot new process
involvement information
7. Eliminate outdated goal(s) from selected sales team members
performance management and create new goal(s) more aligned with
Fail early,
Innovation desired behaviours
learn fast
Example cultural levers Example MVCs for ’collaboration’ element of culture

Adopting an IDO Culture drives data literacy across an organisation – the ultimate goal is then to
reduce an organisation’s change journey
• Organisations will naturally undergo a

performance dip as they undergo transformation
in the analytical space due to new required
learning and resources, resistance to change and
other common blockers that arise
• An IDO Culture aims to reduce this dip. It aims to
enhance the organisation’s level of data literacy
in such a way that:
Performance
Desired performance level

− Users have gained knowledge of emerging
data technologies and understand their
` purpose to value better
Current performance level − Users embrace analytical change and drive it
forwards rather than resist
Reduce length and depth of your performance − D&A projects adopt continuous improvement
dip during a period of change by adopting an in governance, processes and people
IDO Culture
− Users understand and are excited to work
towards the same goal
− Leaders embrace change across their
organisation’s data and analytical capabilities.
Time during organisational change They enable the organisation to drive this
change forward

A pioneering data, AI and analytics culture will manifest itself and its benefits across the 7 key levers
LEADERSHIP STORIES
How leaders think, how this shapes our strategic priorities and What we talk about inside and outside the organisation. Our
impacts all levels vision, strategy, stories, customer insights: our history. Who and
what we choose to immortalise says a lot about what we value
Example behaviour: Leaders develop team members risk
management toolkits to work comfortably with uncertainty and
manage ambiguous situations, then delegate decision-making
processes on them
SYMBOLS INFLUENCE & DECISION-MAKING

Visual and physical representations of the organisation through The way key relationships are developed and how people win
its workplace, including branding, work environment, office
CULTURE support for decision making processes without using position
layout, meeting spaces, and dress codes LANDSCAPE
Example behaviour: People work together seamlessly with
colleagues from all different areas of the organisation
STRUCTURES + INFORMAL NETWORKS PROCESSES & SYSTEMS

The formal and informal relationships for how work is managed The way we run the business. These include financial systems,
and gets done. Defined by both the organisational chart and technology systems, quality systems and rewards (including the
influence or relationship networks way they are measured and distributed) within the organisation
Example behaviour: People use practices and business routines to Example behaviour: People leverage technologies and challenge
communicate effectively with a diverse talent pool; integrating current practices to stays ahead of the competition. Each person
business and industry know-how, with analytics and AI RITUALS + ROUTINES understands the value of sharing knowledge - successes and
Daily rituals and routines of people that signify acceptable failures
behaviour. Sets norms for given situations, defines what’s
valued by management

Organisations should start to see desired benefits across three key areas – Brand and Reputation,
Financial Impacts, Employee and Customer Satisfaction
3X
more profitable and
have up to 4X better
30%
retention in higher levels of 88%
69% organisations that innovation for
companies who tend
increase in employee
of respondents would have a learning satisfaction for one
culture against those to be ‘mission-driven’ company that
not take a job with a
company that had a that do not introduced a
bad reputation recognition program
focused on company
Financial values
Impacts
Employee and
Brand and
Customer
Reputation
31% Satisfaction
84% decrease in voluntary 30%
18% of respondents would
consider leaving their
turnover for
companies that
great customer 40%
of organisations put satisfaction levels for higher level of
current jobs if offered demonstrate high
social responsibility as organisations with an retention for
another role with a levels of recognition
a top priority, yet 77% engaged culture, as companies that are
company that had an
say it’s “important” per data collected ‘mission-driven’
excellent corporate
over a ten-year period
reputation

Top tips for organisations looking to accelerate their IDO Culture
Develop broad data skills - Scale beyond an

Decide with balance - Articulate humans and
analytics team. Demonstrate broad data skills, ask the
machine. Couple data with experience or intuition to
right questions, identify the root cause issues indicated
create impactful decisions that use both the strength of
by data, think holistically about which data might be
the data and human context, ultimately generating the
useful for solving problems, and readily translate and
best outcome for the organisation.
share insights across the organisation.
Demonstrate value - Leverage leaders as role

Determine what works - Design for best fit, not for
models. Build awareness and socialise the value that
best practice. High performing organisations test and
data brings to the organisation. Leaders in high
experiment to determine what works for them. Leverage
performing organisations frequently communicate the
a flexible way of working in order to define a customised
importance of data and treat data as a core asset for
approach that reflects the organisation’s needs.
competitive advantage.

Distributed Enterprise
Businesses are shifting towards remote operating models, driven by the rise of home working
and remote consumers
A DISTRIBUTED ENTERPRISE REQUIRES BUSINESS A DISTRIBUTED ENTERPRISE IS ENABLED THROUGH THE CLOUD
MODELS TO CHANGE AND BENEFITS ARE ENHANCED BY EMBRACING DATA AT THE EDGE*
• Due to the rise of remote and hybrid working patterns, traditional

organisations are evolving into ‘distributed enterprises’ with staff, Analytics and AI
technology, assets and consumers spread widely across the globe
• Organisations must plan for the obvious impact of supporting a
Cloud and Edge orchestration

disparate workforce and the less obvious change of adapting
Workload Management
business models to new customer demands caused by a
distributed enterprise
The
• Improved workload management, data management, networking Distributed
capabilities, computing power, data storage, machine learning and
Enterprise’s Data Management
Technology
automation are key to businesses aspiring to become increasingly
Foundations
distributed
• A leading distributed enterprise is formed on cloud technology, Compute Power and Storage
connected to the edge, driven by digital first principles and
supported by employees who are data literate and technology
enabled
Networking

* see “Analytics on the edge” IDO Playbook 2022 87
Edge and cloud based solutions are a nascent trend that will continue to drive digital transformation
across every industry
CLOUD TECHNOLOGY UNDERPINS THE DISTRIBUTED ENTERPRISE… …WHICH DRIVES THE NEED FOR COMPUTING CLOSE TO THE EDGE*
• As more devices become interconnected through IoT and enterprises become

Hybrid Worker more remotely distributed, large amounts of data is produced at the edge –
or in simpler terms, closer to the data source
• Businesses can benefit from real time low latency analytics by shifting the
Examples Cloud Based Work Technology
computing closer to the edge of the network, outside of the cloud
Digital
Desktop as a Endpoint Collaboration • The number of IoT devices are forecast to triple from 2020 to 2030 growing
Employee
Service Management Enablement at a compound annual growth rate of 11%
Experience
Deliver computer Maintain security Do your best work Work as a team • Manufacturing, Natural Resources, Healthcare Providers and Smart Buildings
services anywhere access to network ` from anywhere from anywhere
are set to be fastest growing segments in IoT
Process Data Software as a • By 2025, more than 50% of enterprise managed data will be created and
Automation Availability Service processed outside of the cloud
Automate Systems and Access software

workflows, focus devices integrated and applications
on adding value to collate and from anywhere
provide access
to data

* see “Analytics on the edge” (slides 67-70) IDO Playbook 2022 88
A distrusted enterprise generates cross-organisational benefits
Scalability
• Cloud technology and collaboration platforms enable you to scale
workforce with ease
• Share and obtain data from the center to the edge of your
organisation
• Drive change and deployment at a faster pace than competition
Revenue
• Interact with new customers and markets
• Revenue growth from digital innovation
• Data monetisation
Consumer Experience
• Cater for the digital consumer
• Wider business reach by opening more C2B/B2C channels
• Reduce friction between consumer and employee
Employee
• Increase productivity through automation
• Access a larger, more diversified talent pool
• Improved retention rates
• Decentralised organisation
Analytical Capability
• Reduce fatigue and burnout
• Collect and analyse more data faster with increased efficiency
• Connect assets to networks, scale predictive analytics and AI
• Become more responsive to change through data – drive innovation
• Transfer useful data to customers more seamlessly
• Deploy advanced, democratised analytical applications
Cost
• Reduced physical infrastructure costs
• Manage and monitor operational costs more seamlessly
• Reduced networking costs

The inevitable shift towards distributed workforce, customers, services and assets means organisations
must be ready to adapt their analytics operating models at pace
SHIFTING TO THE RIGHT ANALYTICS FUNCTION OPERATING MODELS
There are various operating models an organisation can adopt - a ‘one size
fits all’ approach does not apply. Analysts reside in one central group where they serve a variety
Centralised
of functions and business units and work on diverse projects
The scope and capabilities of the function need to be determined, based on Centralised
both the organisation’s current and potential future needs. The size, scale
and level of influence of the function will evolve over time as the business A model which leverages offshore or outsourced capability to
view of AI and thirst for insight matures. Some organisations may transform provide an engine room for analytics which is focused on
from one type of operating model to another as their organisation matures Factory industrialising solutions
and becomes increasingly distributed.
Benefits of the right AI function include: Analysts work together in a central group, but act as internal
• Making better decisions, by joining the dots across business siloes consultants and charge business units for their service
Consulting
• Driving innovation in products, services and internal operations
A central entity coordinates the activities of analysts across
• Improving the speed and reliability and reduce the cost of decision
units throughout the organisation and builds a community to
making Centre of
Excellence share knowledge and best practices
• Enriching the depth of analysis and insight by connecting external data
sources
Analysts are located in functions like marketing and supply
Decentralised
• Improving knowledge sharing, making learning and development easier, chain, where the most analytical activity occur
and democratising across platforms Functional
• Govern the use of AI in a non-pervasive way
Analysts are scattered across the organisation in different
As the distributed enterprise shift continues, organisations need to
functions and business units with little coordination
understand the different operating model approaches for analytical Dispersed
functions and be ready to adapt as necessary.
Organisations should continuously review the structure of their analytical services and the types of
services offered – critical to maintaining a winning position through the distribution hype
ASK SOME CRUNCHY QUESTIONS! EXAMPLE SERVICES
• Which services do we need more Insight Services Enabling Services

investment in? Insight Generation Training Communications Data
• Which services are we missing? Analytics Application Design Analyst Training Capability Communication Master Data Management
• Which services should we start Visualisation User Training Priority Communication Data Extraction
providing? Predictive Modelling Governance Data Availability Communications Data Cleansing
• How are our analytical services Scenario Modelling Demand Management Priority Communications Data Definition
interconnected in a distributed world? Data Privacy & Security
Process Simulation Capability Communications Data Profiling
• Do we need to change our governance Dashboarding Ecosystem Management Service Management Information Model & Data Sources
structure?
Audience Assessment Report Library Maintenance Problem Management Data Governance
• How can we make our data Community of Practice Incident Management Management
management more robust with the
growing scale of data generated across Solution Estate Management Service Desk Portfolio Management
a distributed enterprise? Change Management Cont. Service Improvement Supplier Management
BI Programme Management Event Management Service Level Agreements
Portal Management Financial Management
Production Services Advisory Services

Reports/Dashboards Delivery Support
Exception Reporting Solution Change Cloud Services Toolset Support
Mobile Reports Real Time Location Services Industrialisation Programme Advisory
Real Time Reporting Application Management Access Management Delivery Support
Repeatable Reporting Release & Deployment Management
Dashboard/Report Development Source System Integration Architecture
Dashboard/Report Configuration
Top recommendations for leaders looking to unlock distribution across their organisation
CONSIDER COLLABORATION TOOLS, PIVOT BUSINESS MODELS INVEST IN THE RIGHT TECHNOLOGY TO ENABLE DATA CAPTURE,
AND FOCUS ON DIGITAL TRANSFORMATION REMOTE COLLABORATION AND IMPROVE DATA ANALYTICS
• Reduce employee fatigue by rearchitecting collaboration tools, • Invest in edge computing skills and teams across your data and
workspaces and processes to match new hybrid work environment analytics personnel
• Increase visibility into employee sentiment and experience as • Focus on data management, IoT and processing outside of the cloud
employees adopt modern technology advancements and closer to the edge
• Improve endpoint* performance and IT support • Embrace the Metaverse and Virtual Reality (VR)
• Pivot business models to capture market share from customer • Investment plans should focus on cloud hosted workplace technologies
behaviour changes due to remote working leveraging platform, infrastructure and software as a service, unified
endpoint* management, and desktop as a service (DaaS)
• Adopt virtual first and remote first architectural principles
• 5G coverage and bandwidth across IoT enabled assets – improve
• Provide tools for teams to rapidly develop and improve customer
network connectivity and latency whilst continuing investments in
facing technologies
cybersecurity
• Enhance the experience – remove friction in digital interactions for
• Focus on vendors and tech that can assist in managing on location
both customers and employees
edge computing, shifting AI closer to the edge
• Adapt multi-disciplinary operating models and upskill in data and
• Enable business innovation by reorientating IT teams to collaborate in
digital capabilities. Prepare your ability to be able to pivot analytic
the delivery of self-service platforms
operating models seamlessly through the inevitable shift towards
distributed everything

* Manage employee devices & policies from cloud solution that enables self service IDO Playbook 2022 92
Total Experience
A strategy that creates superior shared experiences by interlinking multi-experience (MX), customer
experience (CX), employee experience (EX) and user experience (UX) disciplines
Many organisations are already working on making experiences better — from the customer, employee, user and multi-experience perspective.
But these are traditionally siloed and business sponsors are all in different departments.
Creating better individual experiences is a great step, but the next step is to interlink and continually refine to create a holistic approach.
Customer Experience Employee Experience

How a customer interacts with and feels How an employee interacts with and feels about their
about a brand CX EX company
• How do we increase engagement? • What is most important?
• How can we keep customers happy? • How do we help them progress?
• What problem can we help them solve? • What can we improve at minimal expense?
Multi-Experience
How an experience is enhanced and delivered User Experience
simultaneously across multiple devices, modalities, How a user interacts with and feels about a product or
and touchpoints experience, particularly within the digital realm
• Do we have the right platforms? MX UX • What are the friction points?
• What provides the right interaction based on • How much effort does it take to get what they need?
each persona? • Does our experience map to the user journey?

Total Experience
While achieving excellence in each individual function is an accomplishment, substantial benefits can be
recognised by organisations that are able to link their activities to one another
1. Breaking down silos & building cross-functional teams 2. Competitive and Comparative Advantage
Data owners in the organisation will almost certainly span across operations, By interlinking CX, MX, UX and EX activities together, organisations make it
marketing, finance, and technology functions. Adopting a total experience more difficult for competitors to imitate the experience they offer. This may
approach to data and analytics will not only interlink the different experiences provide a competitive advantage to the organisation depending on the type of
but also encourage cross-functional collaboration, resulting in efficient data activities and degree to which they are interlinked.
access and management.
3. Efficient Operations 4. Continuous Reinforcement
Having activities linked together instead of operating in silos can result in Linked activities can continuously reinforce and improve each other. For
more efficient operations. For example, an effective information hierarchy example, an organisation that provides an excellent customer experience
used by one experience team may be reused and extended for use by another could apply its CX knowledge to EX by making the CRM tools that employees
team, saving time and resources while ensuring consistency of the experience. use easier to operate. Employees, in turn, become faster and better at serving
customers, resulting in further improvements to CX. The organisation can then
apply the new CX lessons to EX again, starting a new cycle.

Total Experience
Focus from D&A leadership has been towards individual experiences – these only capture a portion of the
optimal operating model
Traditional focus within building data and analytics capabilities has been aimed at producing outcomes from the perspective of a single type of experience.
By incorporating the Deloitte 9 step Target Operating Model Development Framework and adapting it for analytical capabilities, a TX strategy ensures the right leaders
across CX, EX, UX and MX collaborate to identify and remove impediments, uncover new opportunities in their disciplines, and work toward an integrated solution.
Customer and Channels “Understanding the customers of analytics and how they interact” MX
CX
Services “Articulating the services customers need”
Capabilities “Defining the capabilities required to deliver”
Organisation “Building the organisation for analytics delivery”

EX
Talent “Growing the people for analytics delivery”
Location “Positioning the people for optimum delivery”
Processes “Governing the delivery”
Data “Sourcing the data to manage the capability” UX
Technology “Providing the people with the right technology”

Total Experience
Armed with an understanding of what each of these functions are and how to interlink activities among them,
start enabling the functions to reinforce one another
Link to CX Link to MX Link to UX Link to EX
Review the current maturity of customer Determine where unnecessary

Customer Use customer journey maps to
experience and determine how UX can employee effort is preventing
determine which steps among your key
Experience start — or broaden — its role in each employees from providing a better
customer journeys can benefit from MX.
dimension. experience and generating solutions.
Conduct regular consistency checks Deliver frictionless employee

Multi among the various modalities your Build a design system and adopt experiences through effortless customer
Experience organisation offers to customers to DesignOps practices and tools. experiences by shifting from an inside-
ensure information is consistent. out to an outside-in MX perspective.
Create a partnership between UX, CX

Collaborate with marketing/branding to
and marketing to model the
educate the UX team on the brand Create an internal UX ‘academy’ in which
customer acquisition funnel, and
User strategy and identity; ensure UX teams product, CX and marketing employees
identify dependencies and
Experience accurately apply visual, behavioural and learn foundational UX principles,
opportunities within the product to
written guidelines across all relevant MX processes and best practices.
promote adoption and long-term
touchpoints and modalities.
retention.
Establish a multidisciplinary core team

Improve EX by applying the including, but not limited to, IT, business Deploy UX teams to improve EX through
Employee fundamental principles and proven leadership, HR, facilities management, the ‘new work nucleus’, better employee
methods of CX such as employee UX, experience design and product tools for collaboration, and smart
Experience
journey maps, personas and Voice of development to leverage MX physical workspaces, but also through
Employee. approaches to deliver a frictionless work hackathons; bring design thinking to EX.
environment.

What’s next?

What’s next?
The IDO Scaling Lab
IDO scaling labs bring the IDO Methodology to life and accelerates your analytical ambitions. Designed to be fun and interactive, the Labs cultivate collaboration and unity
towards a shared common goal. Utilising a tailored modular approach, the Labs are applicable for organisations from all industries, at any maturity, to expedite and help
you break through barriers to achieve analytics at scale.
AVAILABLE MODULES
Asking the
right questions Explore IDO Target Operating Model
Shared Vision & Ambitions Foster Analytical Culture

Taking the
right actions
Value Capture Apply Digital Ethics
Making the right

impact
Deep Dive Opportunities Ignite Next Steps

What’s next?
The IDO Scaling Lab
What to expect from a lab with us?

Lab
IDO Labs allow you to apply a jump start

to any analytical incentive. It provides a Industry insight, thought leadership and engaging
safe space to be creative and address interactive exercises are designed to help you
organisational barriers to change as well understand the key barriers to scaling; the specific
as aspirations of where you would like to foundational capabilities required to scale; where
be. you can add value; and define actions for IDO
scaling.
To ensure your organisation obtains the
maximum value from having key decision
makers in one space, we support you Analytical
across the entire lifecycle i.e., identifying
the appropriate attendees, through to
Acceleration
actionable next steps.
Post Lab
Pre-Lab Comprehensive output deck that includes an

executive summary, your defined next steps in
relation to each module, a write-up of exercise
Before the Lab, we work closely with you to outcomes and your scaling roadmap for building
understand your AI and Analytics journey to momentum.
date, your key drivers, and the challenges that
are currently holding you back, to tailor for
maximum value.

What’s next?
Experience Analytics
Experience Analytics is Deloitte’s flagship

technology event and our largest client
analytics event in Europe.
Each year, we bring together C-suite, senior executives and
leaders in data, analytics and AI to explore not only key issues
and challenges, but strategies for solving them.
Our agenda covers key themes that will help us better
understand the impact that Data, AI and Analytics are having on
businesses, society, and individuals. Through Ted-X style talks, live
demonstrations and interactive lab sessions, this all-day event will
allow you to engage with thought leaders, learn from your
industry peers and explore the latest insights and trends in data
and AI.
Join us at our next Experience Analytics event to discover
what the power of analytics & AI can do for your
organisation.
https://www.deloitte.co.uk/experience-analytics/
A global presence
We have high quality technical teams across the globe with the capabilities to help deliver digital
transformation to the highest standard and at scale
START A CONVERSATION
21 Global Delivery
Locations
`
47 Analytics Labs and
Digital Studios Globally
Over 54,000 delivery

professionals across the
globe
We deliver the highest quality on a global scale
Our world class global delivery network offers highly experienced and technical support to execute large
complex projects for our clients. Our transformations are executed at scale with quality
Expertise in all aspects of a Digital transformation An Established Analytics Leader

IDC MarketScape: Worldwide Artificial Gartner 2022 Magic Quadrant for Data
CDAO Transition
Insight Strategy & Labs Insight Driven Organisation Intelligence Business Services 2021 and Analytics Service Providers
Business Case Scaling Labs
Simulation and
Digital Twins Portfolio & Programme
Management
Data Science &
Artificial Demand Management
Strategy & Upskilling
Intelligence
Advanced
Intelligent Organise
analytics
Automation Operating Model Design
& Change Management
Deloitte
Analytics
Data Trust &
Governance
Analytics Foundry
Data Deliver
Master Data API Engineering Salesforce Cloud Engineering

Management Data Visualisation & 2021 MuleSoft Global Partner Salesforce Partner Innovation Award Named by IDC as Worldwide Leader
Architecture Rapid Prototyping of the Year & Gartner Top Award for CRM in Industry Cloud Professional
Services
Data Lifecycle
Management
Data & AI Engineering
Full Stack Engineering Data & Analytics
2021 Leader in Gartner MQ for CRM 2022 Leader in Gartner MQ for Data
Data and Application & Customer Experience and Analytics Service Providers
Integration Ecosystem Infrastructure and Implementation Services (14th time) worldwide (6th time)
Architecture Architecture Incident Management
Contacts
UK EMEA
Natalie Williams Linda Smith Bjoern Bringmann
Deloitte UK Deloitte UK Deloitte Germany
Global IDO Lead IDO Scaling Lab Lead [email protected]
[email protected] [email protected]
Andy Gauld Tom Bevan Sander Buijsrogge

Deloitte UK Deloitte UK Deloitte Netherlands
[email protected] [email protected] [email protected]
Varvn Aryacetas Georgie Chapman Hilary Richters

` `
Shalini Gupta Bolly Williams Monika Viktorova
Jane Kwan Loic Liegeois Peder Hinge Pedersen

Deloitte UK Deloitte UK Deloitte Denmark
Kishore Nadarajan Kevin Zhong Francesco Sagrati

Deloitte UK Deloitte UK Deloitte Italy
Appendix
IDO Playbook 2022 – Sources used
Playbook Area Sources
An Insight Driven Scary Smart: The Future of Artificial Intelligence and How You Can Save Our World (Mo Gawdat, 2021),
Organisation Competing on Analytics: The New Science of Winning (Thomas H. Davenport & Jeanne G. Harris, 2017),
Deloitte
Vision Deloitte
Value generation Deloitte
Insight driven performance Deloitte
DIY Analytics Gartner, Dataiku, Deloitte
Digital ethics and AI Deloitte (Netherlands) , Deloitte Trustworthy AI, grandviewresearch.com
Start small, scale fast https://signum.ai/blog/small-and-wide-data-is-important-and-relevant-is-the-era-of-big-data-coming-to-an-end/, Gartner,
https://knowledge.wharton.upenn.edu/article/small-data-new-big-data/, Deloitte
Analytics on the edge Gartner, Deloitte: Our edge computing opportunity SGO Intelligence & Analysis brief, “Sensors and Machine Learning: Glucose Monitoring with An AI Edge,”
AI Trends, October 31, 2019. Microsoft Azure, “Cree uses Azure Data Box Edge to innovate and grow.”
Gina Narcisi, “AT&T, Dell Team Up To Tackle Open-Source Edge Computing, 5G,” CRN, August 15, 2019, Business Wire, “AWS Announces First AWS Local Zone
in Los Angeles,” press release, December 3, 2019, Deloitte
Secure collaboration https://www.forrester.com/report/chief-data-officers-invest-in-your-data-sharing-programs-now/RES164496,
https://www2.deloitte.com/us/en/insights/focus/tech-trends/2022/data-sharing-technologies.html, Deloitte
War for talent https://www.forbes.com/sites/forbeshumanresourcescouncil/2021/06/16/talent-wars-the-post-pandemic-hiring-race-for-a-competitive-
advantage/?sh=592f44895ca4, https://www.hpe.com/us/en/insights/articles/closing-the-data-science-talent-gap-2203.html, Gartner, Deloitte
Adopting IDO culture Deloitte
Distributed enterprise Gartner, HPE Technology, Deloitte
Total experience Gartner, Deloitte
Appendix
This publication has been written in general terms and we recommend that you obtain professional advice before acting or refraining from action on any of the
contents of this publication. Deloitte MCS Limited accepts no liability for any loss occasioned to any person acting or refraining from action as a result of any material
in this publication.
Deloitte MCS Limited is registered in England and Wales with registered number 03311052 and its registered office at 1 New Street Square, London, EC4A 3HQ,
United Kingdom.
Deloitte MCS Limited is a subsidiary of Deloitte LLP, which is the United Kingdom affiliate of Deloitte NSE LLP, a member firm of Deloitte Touche Tohmatsu Limited, a
UK private company limited by guarantee (“DTTL”). DTTL and each of its member firms are legally separate and independent entities. DTTL and Deloitte NSE LLP do
not provide services to clients. Please see www.deloitte.com/about to learn more about our global network of member firms.
© 2022 Deloitte MCS Limited. All rights reserved.


Deloitte Uk Ido Playbook 2022 Final1

Uploaded by

Copyright:

Available Formats

Deloitte Uk Ido Playbook 2022 Final1

Uploaded by

Document Information

Original Title

Copyright

Available Formats

Share this document

Share or Embed Document

Sharing Options

Did you find this document useful?

Is this content inappropriate?

Copyright:

Available Formats

Deloitte Uk Ido Playbook 2022 Final1

Uploaded by

Copyright:

Available Formats

IDO Playbook 2022

Everything we do is focused on enabling fast,

© 2022 Deloitte LLP. All rights reserved. IDO Playbook 2022 2

Natalie Williams Andy Gauld

© 2022 Deloitte LLP. All rights reserved. IDO Playbook 2022 3

Introduction Taking the right actions (cont’d)

5 An Insight Driven Organisation 64 Analytics on the edge

75 War for talent

Taking the right actions What’s next?

44 DIY Analytics 98 IDO Scaling Lab

59 Start small, scale fast 103 Contacts

© 2022 Deloitte LLP. All rights reserved. IDO Playbook 2022 4

© 2022 Deloitte LLP. All rights reserved. IDO Playbook 2022 5

IT’S TIME TO ACT NOW HOW TO SUCCEED?

© 2022 Deloitte LLP. All rights reserved. IDO Playbook 2022 6

ORGANISATIONS THAT CREATE A WINNING

4. Distributed 5. Constant Innovation and 6. Proactive governance

© 2022 Deloitte LLP. All rights reserved. IDO Playbook 2022 7

© 2022 Deloitte LLP. All rights reserved. IDO Playbook 2022 8

Automation with Intelligence

Your business processes with speed and precisions

Collaboration made greater with the machines we invent

© 2022 Deloitte LLP. All rights reserved. IDO Playbook 2022 9

Now we’re telling you AI at scale is next

© 2022 Deloitte LLP. All rights reserved. IDO Playbook 2022 10

SCALE MAXIMISES IMPACT

Many organisations today accept that richer insights delivered in a timely

© 2022 Deloitte LLP. All rights reserved. IDO Playbook 2022 11

How will it help us win and differentiate ourselves Scalable

Innovation & Growth Data & Technology

Operating Model Is the technology reliable? Is it cost effective? Can

© 2022 Deloitte LLP. All rights reserved. IDO Playbook 2022 12

Data driven Insight driven

© 2022 Deloitte LLP. All rights reserved. IDO Playbook 2022 13

What does this mean in practice ?

© 2022 Deloitte LLP. All rights reserved. IDO Playbook 2022 14

© 2022 Deloitte LLP. All rights reserved. IDO Playbook 2022 15

Challenging the status quo

Customisation Re-inventing the wheel rather than exploring the

Analytics is developed in silos and effort is

© 2022 Deloitte LLP. All rights reserved. IDO Playbook 2022 16

© 2022 Deloitte LLP. All rights reserved. IDO Playbook 2022 17

© 2022 Deloitte LLP. All rights reserved. IDO Playbook 2022 18

© 2022 Deloitte LLP. All rights reserved. IDO Playbook 2022 19

The I in IDO now covers the following 5 areas:

Information Insight Intelligence Innovation Industrialisation

© 2022 Deloitte LLP. All rights reserved. IDO Playbook 2022 20

What does good look like?

© 2022 Deloitte LLP. All rights reserved. IDO Playbook 2022 21

What does good look like?

© 2022 Deloitte LLP. All rights reserved. IDO Playbook 2022 22

What does good look like?

What does good look like?

© 2022 Deloitte LLP. All rights reserved. IDO Playbook 2022 24

What does good look like?

© 2022 Deloitte LLP. All rights reserved. IDO Playbook 2022 25

The building blocks

IDO Operating Model

In its entirety, the IDO building blocks