Master's thesis

International Business Management

2014

Irina Pennanen

DATA GOVERNANCE
– Intelligent way of managing data
MASTER'S THESIS | ABSTRACT

TURKU UNIVERSITY OF APPLIED SCIENCES

International Business Management

2014 | 63

Instructors: Laura Heinonen, Matti Kuikka

Irina Pennanen

DATA GOVERNANCE
– Intelligent way of managing data

Today’s business is run by data. Still there are many companies that don’t think data as a business
critical asset. How have we ended up to this situation? World is changing all the time and
companies should change too. What are the benefits of governing data well? How should it be
done in organizations?

This thesis is trying to point out the background for this problem, find out what companies can do
to make situation better and what are the advantages to business that can be achieved with better
data management.

KEYWORDS:

Data, information, data governance, data management

OPINNÄYTETYÖ (YAMK) | TIIVISTELMÄ

TURUN AMMATTIKORKEAKOULU

International Business Management

2014 | 63

Ohjaajat: Laura Heinonen, Matti Kuikka

Irina Pennanen

DATA GOVERNANCE
– Älykäs tapa hallita tietoa
Yritystoiminta perustuu tietoon, joka on nykyään lähes aina sähköisessä, tietokoneiden
ymmärtämässä ja tulkitsemassa muodossa. Koska tietokoneet eivät ajattele kuten ihmiset, on
tiedolle määritettävä muoto ja opetettava tietokone ymmärtämään sitä. Mikäli muoto vaihtelee
sovitusta, tietokoneet ja –järjestelmät eivät toimi.

Yrityksen panostavat koneisiin, rakennuksiin, henkilökuntaan, mutta panostus tiedon laatuun on

suhteellisen vähäistä. Harvassa yrityksessä ymmärretään tiedon laadun merkitys yrityksen
toiminnalle ja vielä harvemmassa on sitä varten organisaatiossa roolit ja määritykset ja tuloksia
seurataan.

ASIASANAT:

Tieto, informaatio, tiedonhallinta, tiedon hallitseminen, tiedon laatu

1 INTRODUCTION – OBJECTIVES OF THE THESIS 6

2 THE HISTORY AND THE FUTURE OF DATA, COMPUTERS AND INFORMATION 8

3 DATA AND INFORMATION 15

4 DATA QUALITY 24

5 DATA GOVERNANCE 35

6 UNDERSTANDING THE IMPORTANCE OF DATA QUALITY IN METSO

AUTOMATION 51

REFERENCES 43

PICTURES

Picture 1. Use of central mainframe. 8

Picture 2. PCjr, The easy one for everyone. 11
Picture 3. Apple invents the personal computer. Again. 12
Picture 4. The fourth dimension. 13

FIGURES

Figure 1. The The Rock (Benson 2012). 16

Figure 2. A simplified taxonomy of data (Benson, 2012). 19
Figure 3. Master data quality management framework. (ISO 8000) 31
Figure 4. Decision domain for data governance. (Khatri&Brown, 2010) 36
Figure 5. The Intelligent Company Model (Marr 2013). 39

TABLE

Table 1. Questionnaire about Data Governance and management issues to Metso

Automation Flow Control executive committee. (2013) 56
LIST OF ABBREVIATIONS (OR) SYMBOLS

OED Oxford English Dictionary

SOA Service Oriented Architecture

SAP client/server enterprise application software

ERP Enterprise Resource Planning –system

PDM Product Data Management –system

DQM Data Quality Management

 6

1 INTRODUCTION – OBJECTIVES OF THE THESIS

Companies are investing money to new factories and machines, to product development,
systems and personnel. But there are only few companies in Finland investing to the
management of data, although companies are run by data.

Why companies should invest to data management? What are the advantages a company
can achieve if doing that? Management is not usually thinking data as a business asset that
it actually is. Especially in smokestack industry data is seen like an enemy: something weird
that management can’t comprehend - complicated IT stuff. It was handled in IT departments
for a long time, but today IT tasks are centralized and more often outsourced and the data
management is transferred under the business organization (where it actually belongs).

Data problems come up usually when company is buying a new system and implementing
that. Good example is ERP (Enterprise Resource Planning) system implementation: prices
of ERP’s are extremely high, and you can only buy the system from the store. But the system
is run by the data, and if the data is wrong, it seems to the management that system is not
working.

I have faced these problems in my career in Metso Automation. I have been involved in
several system projects and ERP roll-outs. All the time the problem is the same: data quality
or actually lack of data quality. In a global company like Metso Automation it is extremely
hard to find out the data owners: who is responsible of certain data content. There are a
huge amount of users who create data to the systems and all the rules and roles should be
clear to everyone. Metso Automation has grown a lot during the years, mostly through
acquisitions. Several companies have combined as a one company and that is the moment
when the problems start with data management. Suddenly there are hundreds of users
creating data instead of tens.

TURKU UNIVERSITY OF APPLIED SCIENCES THESIS | Irina Pennanen

 7

From the system point of view there is of course user rights that defines when someone is
allowed to only view data in the system or also editing it. But because there are hundreds of
editors needed, company should create a governance model for managing the data.
Typically there are a lot of fields in ERP which are mandatory to fill in. For example the
acquisition code: is certain product or part a purchased one or a manufactured one? User
fills the mandatory data but system can’t know if the information is correct or not.

If you don’t have clear rules how to fill the information, you can’t trust it and system is not
working the way people are expecting it to work. Will the rules and roles help then? When
making data governance official it means that management can also monitor the data
quality. If every editor has certain role and responsibility, it can be included to person’s
annual targets: like data correctness must be over 95 percent. When people have targets
and it is measured, it motivates to do things right.

This thesis is trying to open up this issue: not from the IT point of view but from the business
point of view. What are the advantages a company will achieve with governing data and
investing to data management?

TURKU UNIVERSITY OF APPLIED SCIENCES THESIS | Irina Pennanen

 8

2 THE HISTORY AND THE FUTURE OF DATA, COMPUTERS

AND INFORMATION

“Information is the oil of the 21st century, and analytics is the combustion engine,” presented
Mr. Peter Sondergaard, Senior Vice President at Gartner and Global head of Research, in
ITxpo 2011 in Orlando. “Pursuing this strategically will create an unprecedented amount of
information of enormous variety and complexity.”
(http://www.gartner.com/newsroom/id/1824919)

Industrialization started in the end of 1800’s and by the mid of 1950’s it was clear that
automatization is next big thing. So computers started to do some things on behalf of
humans. This required extremely stable data to “teach” the computer to do tasks. Because
of the criticality of data only few people had access to it. This was called centralized data
handling or mainframe usage, see Picture 1:

Picture 1. Use of central mainframe.

TURKU UNIVERSITY OF APPLIED SCIENCES THESIS | Irina Pennanen

 9

2.1 Mainframes

Mainframes were used before microcomputers became general. They were usually very
expensive, powerful and operated by special software. Mainframes were typically used by
large companies, public authorities and universities for their data handling tasks. Typical
tasks done with mainframes:

 File maintenance: This is perhaps the most common use of mainframes. Maintaining
records is a huge task for institutions. Records can contain information on sales, credit
card status, payroll details, social security details, health records, stock inventory, etc.
These either need to be accessed by different people in real-time (for instance a travel
agent booking an airline ticket) or updated in batches (for instance warehouse stock
levels at the end of each day). In such cases it is necessary to have the data stored
centrally and with accessibility for those who need it. A lot of minicomputers are now
capable of performing these tasks in medium-sized companies.

 Simulations: Many physical and engineering problems cannot be solved without the
help of complex computer simulations. These require intensive mathematical work,
and so take advantage of a mainframe's computational power. Examples include
weather forecasting, or calculating the position of astronomical bodies with extreme
accuracy. Many minicomputers or workstations are now used for this type of problem.

 General purpose: Many universities used a mainframe to act as a general purpose

computing facility. Each user can then be given their own area on the mainframe to
store files, and different departments can use its resources to perform different tasks,
e.g. predicting bird populations in the Biology department and calculating metal stress
in the Engineering Department. PCs are now used to perform many of these tasks.
(http://labspace.open.ac.uk/mod/oucontent/view.php?id=426285&printable=1)

TURKU UNIVERSITY OF APPLIED SCIENCES THESIS | Irina Pennanen

 10

Companies used to have own data organizations that only handled data input and also
served rest of the organization to have outputs of the data. This was the way to keep data
quality in high level, 100 % valid. This was working perfect, but of course structures of
organizations were quit heavy.

2.2 Personal computers

Personal computers (PC’s) became more general in 1980’s and centralized model of data
handling were surrendered. Each and every one had their own PC and availability to update
data according the user rights. Unfortunately this has led us to the situation where
company’s business critical information is updated by anyone, without common rules and
follow-up.

The term “personal computer” has been applied to a wide variety of machines (often in
hindsight) where an individual user would have direct control of the entire computer (e.g.,
the LGP-30, Bendy G-15, and others).
(http://www.computerhistory.org/brochures/categories.php?category=thm-42b97f98dbaf2)

However the true personal computer (as we know it today, that is a mass market item found
in both home and office settings) had to await the development of the integrated circuit CPU
in the form of the microprocessor and, more directly, its appearance in machines such as
the Radio Shack TRS-80, the Apple II, and Commodore in 1977. When IBM introduced its
Personal Computer (PC) in 1981, a slow shift in perception began in which the personal
computer changed from being viewed as a toy to a business tool. Today most personal
computers have much greater computational power than even the most powerful
mainframes of only a few decades earlier, Pictures 2 and 3.

(http://www.computerhistory.org/brochures/categories.php?category=thm-42b97f98dbaf2)

TURKU UNIVERSITY OF APPLIED SCIENCES THESIS | Irina Pennanen

 11

Picture 2. PCjr, The easy one for everyone.

TURKU UNIVERSITY OF APPLIED SCIENCES THESIS | Irina Pennanen

 12

Picture 3. Apple invents the personal computer. Again.

In a small company this worked fine: people were sitting next to each other and if the data
creator wrote something wrong the data user asked about it and the data was fixed. But then
companies started to do acquisitions and manufacturing in other locations, meaning that
companies became global. That was the point when the data quality turned out its
significance.

TURKU UNIVERSITY OF APPLIED SCIENCES THESIS | Irina Pennanen

 13

2.3 Business dimensions

Companies have been measuring their business along three dimensions for many years:
people, process and technology. Strout and Eisenhauer have strong opinion that there is a
fourth dimension: data. They believe that data is like blood in the body: its existence is
absolutely necessary to live. By investing heavily in people, process and technology
companies were ensuring their operational efficiencies and gaining insights to their business
and maintaining better control of their business. Thus, data is to operations as blood is to
our body, see picture 4. (Strout&Eisenhauer, 2011, 3)

People

Data

Process Technology

Picture 4. The fourth dimension.

We’ve all heard it for years: “Garbage in, garbage out.” Unfortunately only few businesses
have put in place the appropriate processes to ensure high data quality on a consistent
basis.

Data is a very valuable asset of any business. It is owned by business and it should be
treated as such. This means that it must be protected, guarded, managed and governed in
such a way that it retains or increases in value. Companies spend a lot of money for doing

TURKU UNIVERSITY OF APPLIED SCIENCES THESIS | Irina Pennanen

 14

it other business assets like vehicles, plants, brands, copyrights and patents. John
Eisenhauer has been asking the very proactive question in Data Governance Society
meetings: “If data has an economic value, why shouldn’t we put data on the balance sheet?”
Exactly, should we? We spend a lot of money for creating data, maintaining it and protecting
it. Do we get the most value from it? (Strout&Eisenhauer, 2011, 4)

TURKU UNIVERSITY OF APPLIED SCIENCES THESIS | Irina Pennanen

 15

3 DATA AND INFORMATION

Information versus data… Information is something that gives you the overall description of
the area of interest. One example can be time tables: you can have the data of departing
time: 15.00PM. Okay, that is valid data, but it does not create any information to you. To get
the information you need all pieces of data:

1. the type of vehicle (train)

2. the place of depart (Helsinki, Pasila)
3. the date of depart (1.1.2014).

When you add here the fact that you know what is a train and where Helsinki and Pasila are
located and how to get there, you have the needed knowledge to use that train.

So you may say that information is something you can use to do things. And to get it, you
need lot of data that is valid and only together this data creates useful information for you.

Peter R. Benson, Project Leader for ISO 8000, the International Standard for Data Quality
and the Founding and Executive Director of the Electronic Commerce Code Management
Association (ECCMA) has defined data and information:

“The terms (words) data and information are often used interchangeably as synonyms.
Understanding that they are in fact different concepts is important to understanding data
quality and data governance. You cannot copyright information, only data. Until a song or
a performance is translated into fixed form as data, it cannot be protected under the laws
of copyright. The law of copyright is important to data in many ways, specifically the
concept of a joint work where the work of more than one author is included in a work. Data
is rarely the work of a single author and tracking what is and is not a joint work can be
challenging. (Benson 2012, 23)

TURKU UNIVERSITY OF APPLIED SCIENCES THESIS | Irina Pennanen

 16

Thomas Stearns Eliot (1888-1965) (http://en.wikipedia.org/wiki/T._S._Eliot) wrote a famous

poem “The Rock” that Benson is referring in his book Managing Blind:

“Where is the wisdom we have lost in knowledge?

Where is the knowledge we have lost in information?”

It is common to see the relationship between data and information represented as a pyramid
with data as the base, rising through information and knowledge, to wisdom as the apex:

Wisdom

Knowledge
Justified true belief
(Plato)

Information
Perception of reality
promoted for personal gain

Data
Disruption in a continuum

Figure 1. The The Rock (Benson 2012).

TURKU UNIVERSITY OF APPLIED SCIENCES THESIS | Irina Pennanen

 17

Of course data was conspicuous by its absence in T.S. Eliot’s poem but it is not hard to
understand the omission given that the first “freely programmable computer,” the Z1
Computer, was only invented two years later in 1936 by Konrad Zuse. It was not until the
early 1950s, 25 years later that we saw the first commercial computers.

Because of the close relationship between data and information it is actually very
challenging to find good definitions. According to the international standard for writing
definitions, ISO 704:2009, a good way to test a definition is to substitute it for the term in a
sentence. That is why definitions are written as fragments and not sentences without a
preposition or with a capital at the beginning and a period at the end.
A circular definition is one where when you substitute the term for its definition you end up
with the term used to define itself. (Benson 2012, 24)

Benson has also said that “data is what data is and information is what you make of it”. The
characteristic that determine the quality of data will be inherent to the data itself, while the
characteristics that determine the quality of information require a third party perspective or
opinion. (Benson 2012, 29)

3.1 What is data?

“Data is the raw material for what can become information” states Steven Strout and John
Eisenhauer in their book “The elephant in the room: data”. (Strout&Eisenhauer, 2011, 7)

ISO 8000-2:2012(E) standard defines ‘data’:

symbolic representation of something that depends, in part, on its metadata for

its meaning and other ISO/IEC standard 2382-1:1993, definition 01.01.02:

re-interpretable representation of information in a formalized manner suitable

for communication, interpretation, or processing

TURKU UNIVERSITY OF APPLIED SCIENCES THESIS | Irina Pennanen

 18

Oxford dictionary describes word ‘data’:

 facts and statistics collected together for reference or analysis:

there is very little data available

 the quantities, characters, or symbols on which operations are performed by a

computer, which may be stored and transmitted in the form of electrical signals
and recorded on magnetic, optical, or mechanical recording media.
 Philosophy things known or assumed as facts, making the basis of reasoning or
calculation.

(http://www.oxforddictionaries.com/definition/english/data?q=data)

Peter R. Benson uses the following definitions for a word data:

Data: application processable representation of

or

Data: elements into which information is transformed so that it can be stored or

moved

Data is by its very nature, a historical record. Data is a presentation of entities and events.
Entities are individuals, organizations, locations, goods, services, processes, procedures,
rules and regulations as you can see for the figure 2:

TURKU UNIVERSITY OF APPLIED SCIENCES THESIS | Irina Pennanen

 19

data metadata

transactional
master data
data

identification descriptive classification

property-value-pairs

data data data

Represented by

physical performance data

characteristic characteristic characteristic
s s s

Figure 2. A simplified taxonomy of data (Benson, 2012).

T.S. Eliot’s poem, “where is wisdom we have lost in knowledge and where is the knowledge
we have lost the information,” the answer is simple – it is in the data! Data is what we use
to transfer wisdom, knowledge and information. So data quality and governance are
absolutely critical. (Benson 2012, 27).

Luciano Floridi has defined data in his theoretical article “Semantic Conceptions of
Information” very philosophically:

TURKU UNIVERSITY OF APPLIED SCIENCES THESIS | Irina Pennanen

 20

The Diaphoric Definition of Data (DDD):

A datum is a putative fact regarding some difference or lack of uniformity within some
context.

Depending on philosophical inclinations, DDD can be applied at three levels:

1. data as diaphora de re, that is, as lacks of uniformity in the real world out there. There
is no specific name for such “data in the wild”. A possible suggestion is to refer to
them as dedomena (“data” in Greek; note that our word “data” comes from the Latin
translation of a work by Euclid entitled Dedomena). Dedomena are not to be confused
with environmental data. They are pure data or proto-epistemic data, that is, data
before they are epistemically interpreted. As “fractures in the fabric of being” they can
only be posited as an external anchor of our information, for dedomena are never
accessed or elaborated independently of a level of abstraction. They can be
reconstructed as ontological requirements: they are not epistemically experienced
but their presence is empirically inferred from (and required by) experience. Of
course, no example can be provided, but dedomena are whatever lack of uniformity
in the world is the source of (what looks to information systems like us as) as data,
e.g., a red light against a dark background. Note that the point here is not to argue
for the existence of such pure data in the wild, but to provide a distinction that will
help to clarify why some philosophers have been able to accept the thesis that there
can be no information without data representation while rejecting the thesis that
information requires physical implementation;

2. data as diaphora de signo, that is, lacks of uniformity between (the perception of) at
least two physical states, such as a higher or lower charge in a battery, a variable
electrical signal in a telephone conversation, or the dot and the line in the Morse
alphabet; and

3. data as diaphora de dicto, that is, lacks of uniformity between two symbols, for
example the letters A and B in the Latin alphabet. (Floridi 2013)

TURKU UNIVERSITY OF APPLIED SCIENCES THESIS | Irina Pennanen

 21

3.2. What is information?

Strout and Eisenhauer wrote that “data is the raw material for information”.
(Strout&Eisenhauer, 2011, 7). So that means that information is then the “end product”,
something that has been manufactured from data.

Oxford dictionary defines information:

1. facts provided or learned about something or someone:

a vital piece of information

 [count noun] Law a charge lodged with a magistrates' court: the tenant
may lay an information against his landlord

2. what is conveyed or represented by a particular arrangement or sequence of

things:

genetically transmitted information

 computing data as processed, stored, or transmitted by a computer.

 (in information theory) a mathematical quantity expressing the probability
of occurrence of a particular sequence of symbols, impulses, etc., as
against that of alternative sequences.

(http://www.oxforddictionaries.com/definition/english/information?q=information)

ISO 9000:2005, definition 3.7.1 explains information very simple way:

TURKU UNIVERSITY OF APPLIED SCIENCES THESIS | Irina Pennanen

 22

Information is meaningful data.

ISO/IEC 2382-1:1993, definition 01.01.01 describes information:

Knowledge concerning objects, such as facts, events, things, processes, or

ideas, including concepts that within a certain context has a particular meaning.

Luciano Floridi has also defined information in his theoretical article “Semantic Conceptions
of Information”:

“It is common to think of information as consisting of data. It certainly helps, if only to a

limited extent. For, unfortunately, the nature of data is not well-understood philosophically
either, despite the fact that some important past debates — such as the one on the given
and the one on sense data — have provided at least some initial insights. There still remains
the advantage, however, that the concept of data is less rich, obscure and slippery than that
of information, and hence easier to handle. So a data-based definition of information seems
to be a good starting point.

Over the last three decades, several analyses in Information Science, in Information
Systems Theory, Methodology, Analysis and Design, in Information (Systems)
Management, in Database Design and in Decision Theory have adopted a General
Definition of Information (GDI) in terms of data + meaning. (GDI has become an operational
standard, especially in fields that treat data and information as reified entities (consider, for
example, the now common expressions “data mining” and “information management”).
Recently, GDI has begun to influence the philosophy of computing and information.

A clear way of formulating GDI is as a tripartite definition:

TURKU UNIVERSITY OF APPLIED SCIENCES THESIS | Irina Pennanen

 23

The General Definition of Information (GDI):

σ is an instance of information, understood as semantic content, if and only if:

(GDI.1) σ consists of one or more data;

(GDI.2) the data in σ are well-formed;

(GDI.3) the well-formed data in σ are meaningful.

GDI requires a definition of data. According to (GDI.1), data are the stuff of which information
is made. We shall see that things can soon get more complicated. (Floridi 2013)

So as a conclusion we can say that information is pieces of data collected together and
interpreted by human being. In companies there are a lot of data: cost centers, accounts,
routings, machines, items, documents, products, parts, persons, salaries. To get the facts
how company is doing, you need not just collect, but connect all this data together to get the
information needed. This makes all data critical: if you have exactly correct time for depart
but you are in a bus station instead of a train station that data is worth of nothing.

TURKU UNIVERSITY OF APPLIED SCIENCES THESIS | Irina Pennanen

 24

4 DATA QUALITY

How to define data quality? Quality is probably one of the most misunderstood concepts.
We all know what quality is, yet we cannot really define it. (Benson 2012, 36)

ISO 9000, the standard that contains the terminology for the 9000 series of standards
contains a definition for the term quality:

“degree to which a set of inherent characteristics fulfills requirements”

This introduces two very important concepts; first the “degree to which”, something we can
measure and secondly, the most important “fulfills requirements.”

Quality is about “fulfilling requirements.” We cannot measure data quality unless we can
specify the requirements for data.

The OED defines “requirement” as: a thing that is needed or wanted or a thing that is
compulsory; a necessary condition. The definition in ISO 9000 simply adds that the need or
expectation must be “stated” and this is important.

ISO 9001 is the standard that contains the clauses that you must comply with if you wish to
be ISO 9001 compliant. The difference between ISO 9001 and ISO 8000 is how the
requirement is “stated” and how compliance is measured. To be compliant with ISO 9001
you must have “documented” the requirements. ISO 8000 takes this one step further. ISO
8000 mandates that requirements must be “stated” in a computer processable form. This is
critical to ISO 8000; after all it is a standard about computer processable data, so it makes
sense that the compliance with requirements must be capable of being performed by a
computer and not by a person with a clip board. (Benson 2012, 38-39)

TURKU UNIVERSITY OF APPLIED SCIENCES THESIS | Irina Pennanen

 25

Oxford English Dictionaries

(http://www.oxforddictionaries.com/definition/english/quality?q=quality) defines quality:

noun (plural qualities)

o [mass noun] the standard of something as measured against other things of a

similar kind; the degree of excellence of something: an improvement in product
quality [count noun]:people today enjoy a better quality of life
o general excellence of standard or level: a masterpiece for connoisseurs of
quality
o (usually qualities) British short for quality paper.
o archaic high social standing: commanding the admiration of people of quality
o [treated as plural] archaic people of high social standing: he’s dazed at being
called on to speak before quality
o distinctive attribute or characteristic possessed by someone or something: he
shows strong leadership qualities the plant’s aphrodisiac qualities
o Phonetics the distinguishing characteristic or characteristics of a speech
sound.
o Astrology any of three properties (cardinal, fixed, or mutable), representing
types of movement, that a zodiacal sign can possess.

adjective
o informal
o of good quality; excellent: he’s a quality player

Data and information quality are now widely recognized problems in companies large and
small, ranging from manufacturing and processing, to finance and health care. Incomplete
or duplicate records, poor quality descriptions and inaccurate information cause inefficient
allocation and use of resources. This can add up to a 20% increase to direct and indirect
costs. Poor quality data is a barrier to effective marketing and the leading cause of

TURKU UNIVERSITY OF APPLIED SCIENCES THESIS | Irina Pennanen

 26

transparency issues that drive up the cost of regulatory compliance.

(http://www.eccma.org/iso8000/iso8000home.php)

Already Albert Einstein (1879-1955) (http://einstein.biz/biography.php) understood the

problem when he defined insanity as “doing the same thing over and over again and
expecting different results.” Solving data quality and governance issues requires change,
and changing a stable system requires effort and effort requires motivation. Individuals are
motivated by greed and fear. In order to address the issues of data quality and governance
we must be clear about our motives, our goals and our objectives. We must also find ways
to translate these in ways that will motivate others. (Benson 2012, 5).

From a quality perspective, only two moments matter in a piece of data’s lifetime: the
moment it is created and the moment it is used. The quality of data is fixed at the moment
of creation. But we don’t actually judge that quality until the moment of use. If the quality is
deemed to be poor, people typically react by working around the data or correcting errors
themselves. But improving data quality isn’t about heroically fixing someone else’s bad data.
It is about getting the creators of data to partner with the users — their “customers” — so
that they can identify the root causes of errors and come up with ways to improve quality
going forward. (http://hbr.org/2013/12/datas-credibility-problem/ar/1)

The good news is that a little communication goes a very long way. Time and time again, in
meetings with data creators and data users, I’ve heard “We didn’t know that anyone used
that data set, so we didn’t spend much time on it. Now that we know it’s important, we’ll work
hard to get you exactly what you need.” Making sure that creators know how data will be
used is one of the easiest and most effective ways of improving quality.

Even better news is that addressing the vast majority of data quality issues does not require
big investments in new technologies or process reengineering. To be sure, disciplined
measurement, automated controls, and methodologies like Six Sigma are helpful,
particularly on more sophisticated problems, but the decisive first step is simply getting users
and creators of data to talk to each other. (http://hbr.org/2013/12/datas-credibility-
problem/ar/2)

TURKU UNIVERSITY OF APPLIED SCIENCES THESIS | Irina Pennanen

 27

Once a company realizes that its data quality is below par, its first reaction is typically to
launch a massive effort to clean up the existing bad data. A better approach is to focus on
improving the way new data are created, by identifying and eliminating the root causes of
error. Once that work has been accomplished, limited cleanups may be required, but
ongoing cleanup will not. (http://hbr.org/2013/12/datas-credibility-problem/ar/2)

Very often, data creators are not linked organizationally to data users. Finance creates data
about performance against quarterly goals, for example, without considering how Sales will
want to use them or Customer Service analyzes complaints but fails to look for patterns that
would be important to product managers.

When quality problems become pervasive or severe, the organizational response is often to
task the IT department with fixing them, usually by creating a special unit in the group to
spearhead the initiative. This may seem logical, since IT is a function that spans all silos.
But IT departments typically have little success leading data quality programs. That’s
because data quality is fixed at the moment of creation. With rare exceptions, that moment
does not occur in IT. To address problems, IT people can talk to creators and users, but
they can’t change the offending business processes. All they can do is find and correct
errors, which, as we’ve seen, is not a long-term solution. (http://hbr.org/2013/12/datas-
credibility-problem/ar/3)

Smart companies place responsibility for data quality not with IT but with data creators and
their internal data customers. For most companies, the real barriers to improving data quality
are that some managers refuse to admit their data aren’t good enough, and others simply
don’t know how to fix poor-quality data. The first bit of progress occurs when a manager
somewhere in the organization (possibly a senior executive, but more often someone in the
middle) gets fed up and decides that “there has to be a better way.” The manager launches
a data program and, if the prescriptions noted here are followed, usually gets good results.
(http://hbr.org/2013/12/datas-credibility-problem/ar/3)

TURKU UNIVERSITY OF APPLIED SCIENCES THESIS | Irina Pennanen

 28

4.1 ISO 8000 – The international standard for data quality

A standard is a document that provides requirements, specifications, guidelines or

characteristics that can be used consistently to ensure that materials, products, processes
and services are fit for their purpose. (http://www.iso.org/iso/home/standards.htm)

Standard are developed by committees and they represent negotiated compromises

between domain experts. Standards therefore represent the consensus of domain experts.
Under the ISO procedure a standard must go through a series of ballots. An important part
of the ISO process is ballot comment resolution. A committee must address and answer all
the comments raised by member of countries through their national Technical Advisors
Group (TAG). In the end, a two thirds majority is required to publish an international
standard. (Benson 2012, 29)

What are the benefits of using standards? Standards ensures that products and services
are safe, reliable and of good quality. For business, they are strategic tools that reduce costs
by minimizing waste and errors, and increasing productivity. They help companies to access
new markets, level the playing field for developing countries and facilitate free and fair global
trade. (http://www.iso.org/iso/home/standards.htm)

International standards bring technological, economic and societal benefits. They help to
harmonize technical specifications of products and services making industry more efficient
and breaking down barriers to international trade. Conformity to International Standards
helps reassure consumers that products are safe, efficient and good for the environment.
(http://www.iso.org/iso/home/standards/benefitsofstandards.htm)

TURKU UNIVERSITY OF APPLIED SCIENCES THESIS | Irina Pennanen

 29

In ISO 8000 development the committee identified five characteristics of data that determine
data quality. These characteristics are:

 syntax
 semantic encoding
 meets requirements
 provenance
 accuracy
 completeness

(Benson 2012, 28-29)

In ISO 8000-1:2011(E) standard defines:

 data quality involves data being fit for purpose, i.e. decision it is used in
 data quality involves having the right data in the right place at the right
time
 data quality involves meeting agreed customer data requirements
 data quality involves preventing the recurrence of data defects by
improving processes to prevent repetition and eliminate waste. (ISO
8000-1:2011(E), 6 Principles of data quality)

4.2 Data quality management

ISO 8000-150 Data quality – Master data: Quality management framework describes:

“The ability to create, collect, store, maintain, transfer, process and present data to support
business processes in a timely and cost both an understanding of the characteristics of the
data that determine its quality, and an ability to measure, manage and report on data quality.”
(ISO 8000-150:2011(E), 6) effective manner requires

TURKU UNIVERSITY OF APPLIED SCIENCES THESIS | Irina Pennanen

 30

Enterprises need Data Quality Management (DQM) to respond to strategic and operational
challenges demanding high-quality corporate data. Hitherto, companies have mostly
assigned accountabilities for DQM to Information Technology (IT) departments. They have
thereby neglected the organizational issues critical to successful DQM. With data
governance, however, companies may implement corporate-wide accountabilities for DQM
that encompass professionals from business and IT departments. (Weber et al. 2009)

To manage the master data quality successfully, organizations shall keep the following
fundamental principles:

 Involvement of people: people at all levels who have roles for data quality
management are involved to improve data quality of an organization. Although data
processing of end users with lower-level role has the most direct effect on data
quality, intervention or control of data administrators with middle-level role is required
to implement and settle down processes for data quality improvement in the
organization. In addition, involvement of managers who are in charge of organization-
wide data quality with high-level role is inevitable to change and optimize roles,
authority, and processes of the organization.

 Process approach: data-centric measurement and correction is not enough to

improve data quality of the whole organization. Desired data quality is achieved more
efficiently when activities and related resources for data quality are managed by
processes.

 Continual improvement: data quality is improved continuously through the processes

of data processing, data quality measurement and data error correction. However,
with these processes only, identical data errors that occur repeatedly cannot be
prevented. Recurrence of data errors
can be prevented when the processes to analyze, trace and improve root causes
which hinder data quality goes with these processes. For this, management
processes concerned with data architecture/schema, data stewardship and data flow
shall also be supported. In addition,

TURKU UNIVERSITY OF APPLIED SCIENCES THESIS | Irina Pennanen

 31

organizations shall improve not only processes for data quality management but also
business processes where data are directly operated.

 Master data exchange: all processes to manage master data quality comply with
requirements that can be checked by computer for the exchange, between
organizations and systems, of master data that consists of characteristic data.

Master data quality management framework is built for three processes and three roles as
can be seen in figure 3:

Data Data quality Data quality

operations monitoring improvement

Data
Data
Data Data quality stewardship
architecture
manager planning /flow
management management

Data Data error

Data Data quality
administrator cause
design criteria setup
analysis

Data Data Data quality Data error

technician processing measurement correction

= role = process

Figure 3. Master data quality management framework. (ISO 8000)

TURKU UNIVERSITY OF APPLIED SCIENCES THESIS | Irina Pennanen

 32

4.2.1 Three data quality management processes

The three top-level processes in the framework shall be:

 data operations
 data quality monitoring
 data quality improvement

The data operations process identifies factors that affect data quality and ensures data is
available at the right place in a timely manner. This top-level process shall consist of the
following processes:

 data architecture management; the process that manages organization-wide data

architecture from the integrated perspective to use data in distributed information
systems with consistency and therefore ensure data quality
 data design; the process that designs data schema, and implements a database to
make data users apply data without mistake and ensure data quality
 data processing; the process that creates, searches, updates, deletes data in
accordance with guidelines of data operations

The data quality monitoring process identifies data errors through a systematic approach.
This top level process shall consist of the following processes:

 data quality planning; the process that sets up objectives of data quality in alignment
with the strategies of an organization, identifies factors to be managed, and performs
actions in order to accomplish objectives. This process also includes assurance of
data quality and adjustment of objectives on the back of assurance results
 data quality criteria setup; the process that sets criteria that include characteristics of
data, target data, and methods to measure
 data quality measurement; the process that measures target data with the criteria set
in the process of data quality criteria setup on a real time basis or periodically

TURKU UNIVERSITY OF APPLIED SCIENCES THESIS | Irina Pennanen

 33

The data quality improvement process corrects data errors detected and eliminates root
causes of the data errors by tracing and identifying them. In order to support the top-level
process effectively, adjustment of data stewardship in accordance with data flows tracing is
required. This process has the function of process improvement not only data quality
improvement. Processes for data management are improved at the data administrator level
while business processes at the data manager level. This top-level process shall consist of
the following processes:

 data stewardship and flow management; the process that analyses data operations
and data flows among businesses or organizations, identifies responsible parties and
their data operation systems which influence data quality, and manages the
stewardship of data operations
 data error cause analysis; the process that analyses root causes of data errors and
prevents a recurrence of the same errors fundamentally
 data error correction; the process that corrects the data that turns out erroneous

4.2.2 Three data quality management roles

The three roles in the framework are responsible for performing the processes in the
framework. These roles shall be:

 data manager
 data administrator
 data technician
The data manager shall perform the following processes within the framework:

 data architecture management

 data quality planning
 data stewardship and flow management

TURKU UNIVERSITY OF APPLIED SCIENCES THESIS | Irina Pennanen

 34

The data manager performs the role that directs a guideline for master data quality
management in compliance with objectives of an organization, manages factors that impact
data quality at an organization level, and establishes the plans for performing data quality
activities in the organization. Along with each major top-level process, the data manager
maintains data consistency in individual information systems through the organization-wide
data architecture management, and analyzes factors that affect data quality in data quality
planning. In addition, the data manager takes a role of granting data administrator’s authority
to trace and correct data over the information systems or organization.

TURKU UNIVERSITY OF APPLIED SCIENCES THESIS | Irina Pennanen

 35

5 DATA GOVERNANCE

Data management is defined as the act of managing data, while data governance is
concerned with managing the rules for managing data, the “authority” over the data. The
concept of data governance is still new so what is included and what is excluded varies from
one company to another.

It is important to remember that governance is first and foremost about people. Authority
implies a chain of command and the delegation of duties and responsibilities.

The meaning of the word ‘governance’ according to Institute of Governance:

“The need for governance exists anytime a group of people come together to accomplish an
end. Though the governance literature proposes several definitions, most rest on three
dimensions: authority, decision-making, and accountability. Shortly:

Governance determines who has power, who makes decisions, how other players make
their voice heard and how account is rendered.” (http://iog.ca/defining-governance/)

In article “Designing Data Governance” Khatri & Brown are defining governance:

“Governance refers to what decisions must be made to ensure effective management and
use of IT (decision domains) and who makes the decisions
(locus of accountability for decision making).” (Khatri&Brown, 148. 2010)

Governance in general “refers to the way the organization goes about ensuring
that strategies are set, monitored, and achieved” (Rau 2004, 35 in Weber et al. 2009).

In successful data governance there are several decision domains that should be defined
as seen is figure 4 below:

TURKU UNIVERSITY OF APPLIED SCIENCES THESIS | Irina Pennanen

 36

Data principles

Metadata
Data quality Data lifecycle
Data access

Figure 4. Decision domain for data governance. (Khatri&Brown, 2010)

If summarizing governance and data definitions it makes sense that a company needs to
take care of the data quality and it is possible only if there is a governing model in use.

One of the primary functions of a data governance program is to identify the sources and
the applications of data. This can be a very challenging mission. It is not uncommon for
businesses to be buying data from many sources only to find that not only are the sources
duplicative, but worse they are conflicting. Defining the rules for resolving these conflicts is
an important part of data governance. (Benson 2012, 66)

By answering to questions: who, what, when, where, why and how it will be quite clear what
Data Governance means.

Who?

 Data Governance is of concern to any individual or group who has an interest in how
data is created, collected, processed and manipulated, stored, made available for
use, or retired. We call such people Data Stakeholders

What?

 Data Governance means "the exercise of decision-making and authority for data-
related matters."

TURKU UNIVERSITY OF APPLIED SCIENCES THESIS | Irina Pennanen

 37

 More specifically, Data Governance is "a system of decision rights and

accountabilities for information-related processes, executed according to agreed-
upon models which describe who can take what actions with what information, and
when, under what circumstances, using what methods."

When?

 Organizations need to move from information governance to formal Data Governance

when one of four situations occur:

o The organization gets so large that traditional management isn't able to

address data-related cross-functional activities
o The organization's data systems get so complicated that traditional
management isn't able to address data-related cross-functional activities.
o The organization's Data Architects, SOA teams, or other horizontally-focused
groups need the support of a cross-functional program that takes an enterprise
(rather than siloed) view of data concerns and choices.
o Regulation, compliance, or contractual requirements call for formal Data
Governance.

Where?

 Data Governance can be placed within Business Operations, IT, Compliance/Privacy, or

Data Management organizational structures. What's important is that they received
appropriate levels of leadership support and appropriate levels of involvement from Data
Stakeholder groups.

Why?

 Data Governance Frameworks help us organize how we think and communicate about
complicated or ambiguous concepts. The use of a formal framework can help Data

TURKU UNIVERSITY OF APPLIED SCIENCES THESIS | Irina Pennanen

 38

Stakeholders from Business, IT, Data Management, Compliance, and other disciplines
come together to achieve clarity of thought and purpose.

The use of a framework can help management and staff to make good decisions - decisions
that stick. It can help them reach consensus on how to "decide how to decide." That way,
they can more efficiently create rules, ensure that the rules are being followed, and to deal
with noncompliance, ambiguities, and issues.

How?

 Organization decides what's important to them - what their program will focus on. Then
they agree on a value statement for their efforts. This will help establish scope and to
establish goals, success measures, and metrics.
 Next, develop a roadmap for the efforts, and use this to acquire the support of
stakeholders. Once achieved, design a program, deploy the program, go about the
processes involved in governing data, and perform the processes involved in monitoring,
measuring, and reporting status of the data, program, and projects.

(http://www.datagovernance.com/adg_data_governance_basics.html)

5.1 Benefits of good quality data

Managers should make the right decision every time they make a decision. This brings with
it the need for correct and adequate information. A decision should be based on real
knowledge, which is a holistic point of view and not on “educated guesses”, intuitive feelings
or limited information that is only looked at from some point of view. (Reunanen 2013)

Bernand Marr, Founder and CEO, Advanced Performance Institute, has introduced an
intelligent company model for management: how to make good decisions based on agreed
requirements, collected and qualified data and analyzes of the data. Figure 5 Intelligent
company model (Marr 2013):

TURKU UNIVERSITY OF APPLIED SCIENCES THESIS | Irina Pennanen

 39

The Intelligent Company Model

1.
Stragetic review
Agree information
needs

5. 2.
Make better Collect the right
decisions data
Information and
analytics
infrastructure

4. 3.
Communicate the Analyze data to
insights extract insights

Figure 5. The Intelligent Company Model (Marr 2013).

Marr also makes seven conclusions about using the Intelligent Company Model:

1. Making better decisions is everyone’s everyday job

2. Intelligent Strategy: Clear objectives, strategy maps and questions
3. Intelligent Data: Leveraging big and small data to answer your questions
4. Intelligent Insights : Analytics and experiments
5. Intelligent Communication: Balance numbers with narratives & visuals
6. Intelligent Decision Making: Court-style decision meetings
7. Create a culture of fact-based decision-making

(Delivering Business Insights From Analytics and Big Data, Marr 2013)

TURKU UNIVERSITY OF APPLIED SCIENCES THESIS | Irina Pennanen

 40

To be able to do the right decisions needed, managers should have the right data in use
and available. When data are unreliable, managers quickly lose faith in them and fall back
on their intuition to make decisions, steer their companies, and implement strategy. They
are, for example, much more apt to reject important, counterintuitive implications that
emerge from big data analyses. (http://hbr.org/2013/12/datas-credibility-problem/ar/1)

Fifty years after the expression “garbage in, garbage out” was coined, we still struggle with
data quality. But I believe that fixing the problem is not as hard as many might think. The
solution is not better technology: it’s better communication between the creators of data and
the users of the data; a focus on looking forward; and, above all, a shift in responsibility for
data quality away from IT folks, who don’t own the business processes that create the data,
and into the hands of managers, who are highly invested in getting the data right.
(http://hbr.org/2013/12/datas-credibility-problem/ar/1)

5.2 Meaning of qualified data in business decisions

International Standards are strategic tools and guidelines to help companies tackle some of
the most demanding challenges of modern business. They ensure that business operations
are as efficient as possible, increase productivity and help companies access new markets.

Benefits include:

 Cost savings - International Standards help optimize operations and therefore

improve the bottom line
 Enhanced customer satisfaction - International Standards help improve quality,
enhance customer satisfaction and increase sales
 Access to new markets - International Standards help prevent trade barriers and
open up global markets
 Increased market share - International Standards help increase productivity and
competitive advantage

TURKU UNIVERSITY OF APPLIED SCIENCES THESIS | Irina Pennanen

 41

 Environmental benefits - International Standards help reduce negative impacts

on the environment

Businesses also benefit from taking part in the standard development process.
(http://www.iso.org/iso/home/standards/benefitsofstandards.htm)

Levelling the playing field: How companies use data to create advantage is an Economist
Intelligence Unit report, sponsored by SAP. The Economist Intelligence Unit conducted the
survey and analysis and wrote the report.
(http://www.economistinsights.com/technology-innovation/analysis/levelling-playing-field,
10.12.2013)

The report’s quantitative findings come from a survey of 602 senior executives, conducted
in September 2010. The Economist Intelligence Unit’s editorial team designed the survey.
(http://fm.sap.com/images/kern/assets/sap_EIU_Levelling_The_Playing_Field.pdf,
10.12.2013)

Nearly all companies realize that the way to gain a competitive advantage is to obtain better
data, interpret them quickly, and distribute them in easier-to-use formats. However, there
are many obstacles to the effective use of data and few companies surmount them all—a
fact that results in a lot of unused corporate data. Indeed, only 17% of companies use 75%
or more of the data they collect.
How are companies using information to beat their rivals and create a more level playing
field?

Below is a list of its major findings:

 Leading companies are keenly focused on data. Of the 38% of respondents who say
their company performs ahead of its peers, 74% say that data are “extremely
valuable” in achieving competitive advantage. The best corporate users of data
devote substantial time to figuring out what sort of information they should track and

TURKU UNIVERSITY OF APPLIED SCIENCES THESIS | Irina Pennanen

 42

who within their companies needs it. They also invest in technology and training to
make sure individual workers are able to capitalize on the data they have collected.

 Accuracy trumps detail. Accuracy and timeliness are the most important attributes of
data, ahead of the amount of detail the data offers. This is because getting the basic
insight—about a new prospect, a change in the price of some raw material, or an
emerging problem at a manufacturing plant—is more important than being able to
analyze every detail about it.

 Information supports competition in myriad ways. Seventy-seven percent of

respondents say data make an important contribution to their customer
support/customer relations efforts, and 71% say it helps them support their sales
processes. Operations, cost management and product development are all aided by
data as well. A less common benefit—cited by around half of all companies—is the
contribution that business insights have made to helping executives strengthen
awareness of a company’s brand.

 Yet most companies remain awash in unused data. In fact, only 27% of respondents
say their firms do a better job of using information than most of their competitors. A
large amount of data sitting on a company’s servers, unused, is not uncommon and
can be a sign of a sub-optimal data strategy. In some cases, however, there are good
reasons to hold on to older data. Financial service firms often need archived data as
a defense against litigation; others may want data for future data-mining purposes.

 A top-down approach may stifle competitiveness. Companies sometimes end up

unintentionally approaching data from a management perspective and ignoring its
value to others lower down the hierarchy. The companies that find ways to
“democratize” their data often gain an advantage. Indeed, 77% of the companies that
aim their data initiatives at all employees, regardless of level, say they’ve found ways
to make data extremely valuable to their business. Only 65% of companies where the
data initiatives are intended primarily for managers agree.

TURKU UNIVERSITY OF APPLIED SCIENCES THESIS | Irina Pennanen

 43

REFERENCES

Benson, Peter R., Managing Blind, 2013

http://cacm.acm.org/magazines/2010/1/55771-designing-data-governance/

http://www.computerhistory.org/brochures/categories.php?category=thm-42b97f98dbaf2,
9.12.2013

http://www.datagovernance.com/adg_data_governance_basics.html, 10.12.2013

http://www.eccma.org/iso8000/iso8000home.php, 10.12.2013

http://www.economistinsights.com/technology-innovation/analysis/levelling-playing-field,
10.12.2013

http://einstein.biz/biography.php, 10.12.2013

Floridi, Luciano, "Semantic Conceptions of Information", The Stanford Encyclopedia of

Philosophy (Spring 2013 Edition), Edward N. Zalta (ed.), URL =
<http://plato.stanford.edu/archives/spr2013/entries/information-semantic/>.

http://www.gartner.com/newsroom/id/1824919, 2.11.2013

http://hbr.org/2013/12/datas-credibility-problem/ar/1, 9.12.2013

http://iog.ca/defining-governance/, 25.3.2014

http://www.iso.org/iso/home/standards.htm, 10.12.2103

http://www.iso.org/iso/home/standards/benefitsofstandards.htm, 10.12.2103

http://labspace.open.ac.uk/mod/oucontent/view.php?id=426285&printable=1, 9.12.2013

Liker, Jeffrey K. The Toyota Way, 2010

Marr, Bernard, Intelligent Business Model, 2013

http://www.oxforddictionaries.com/definition/english/data?q=data, 10.12.2103

http://www.oxforddictionaries.com/definition/english/information?q=information, 10.12.2103

TURKU UNIVERSITY OF APPLIED SCIENCES THESIS | Irina Pennanen

 44

http://www.oxforddictionaries.com/definition/english/quality?q=quality, 10.12.2013

Reunanen, Tero, Leader’s Conscious Experience Towards Time, 2013

http://fm.sap.com/images/kern/assets/sap_EIU_Levelling_The_Playing_Field.pdf,
10.12.2013
Strout, Steven; Eisenhauer, John. The Elephant in the Room: Data. What you need to know
to best govern and manage your enterprise data. 2011

TURKU UNIVERSITY OF APPLIED SCIENCES THESIS | Irina Pennanen