Creating great

customer and
employee experiences
with Virtual Agent
Practical guidance for a
successful implementation

START
1
Meeting the expectations
Customers today have increasingly high expectations—their
questions should be answered easily, quickly, and at any time. But
businesses struggle to meet these demands. Support agents are
overloaded, continuously handling simple, common requests that
could have been resolved through the neglected knowledge bases
and support portals.

ServiceNow Virtual Agent—an enterprise chatbot solution—

empowers customers, partners, and employees with consistent,
scalable, and always-on self-service interactions. Virtual Agent
improves service quality and response times by delivering
instantaneous responses 24/7, reduces agent workload by
automating routine responses, and boosts customer satisfaction
through personalized, contextual conversations.

#1
Prepare
#2
Design
#3
Build
#4
Launch

2
Preparing to create a
Virtual Agent
Set the right expectations
Virtual Agent will have the most impact on your business when
used to automatically resolve a high volume of routine issues.
Your employees and customers already have digital personal
assistants such as Siri, Alexa, and Google Home in their lives. But
trying to replicate those abilities in Virtual Agent is unlikely to
have a high return on investment. Focus your efforts on building
specific workflows that will improve business efficiency and increase
customer satisfaction.

Understand required resources and skillsets

Many Virtual Agent conversations can be built by non-technical
roles like a business analyst, but complex topics that kick off
workflows across multiple systems will usually require developers
and scripting. Implementation teams should include a mix of these
resources:

• Business analysts who gather use cases and map

conversation flows

• Service/Process owners who understand the end-to-end flow of a

process and can identify success metrics

• Business owner: This role helps define the decision trees, sets the
priority for conversations to be built, and is responsible for defining
the vision. Depending on the organization, the service or process
owner may also take on these responsibilities.

• Virtual Agent developers who are comfortable with JavaScript

coding and using the Virtual Agent Designer

• Executive sponsors who provide organizational support for

the project
3
Create a governance plan

As you’re assembling the implementation team, create a

document that clearly explains the requirements and procedures
for identifying, selecting, and implementing Virtual Agent topics
in your organization. This step may seem premature early in the
deployment, but will simplify future requests from business units
and stakeholders as Virtual Agent adoption increases. A key part
of this step is to define how you will monitor and improve the
quality of conversations over time.

Identify success metrics

You’ll also need to spend time establishing how to measure the
success of your deployment efforts. There are two aspects of
this to track: the adoption and quality of the Virtual Agent, and
the business impact on cost savings and customer satisfaction.
Each should have appropriate targets defined, and ServiceNow
Performance Analytics can be used to track and visualize your
progress towards these goals.

Adoption of the Virtual Agent can be measured with KPIs such as

active users, new users, or conversation sessions, and the quality
of the user experience can be captured through user experience
survey ratings or metrics such as average handling time and the
number of transfers to live agents. These KPIs will serve as leading
indicators for the cost savings achieved through high deflection
and self-service rates. They also provide early insight into
customer satisfaction (CSAT), customer effort score (CES), and net
promoter score (NPS).

4
Designing a Virtual Agent
Discover use cases and select topics
The first step in the design phase is to spend time studying
your customers, employees, and organizational data to define
and document potential use cases. You can uncover these
topic ideas through user interviews, by meeting with agents to
discuss their most redundant tasks, and by analyzing existing
data about the most common types of incidents or the most
actively used knowledge articles. You can then prioritize the list
of use cases based on business impact and select a subset of
conversations for development.

Create a unique experience and a personality

A key benefit of Virtual Agent is the ability to access and use
contextual information such as the user’s profile, location,
or items they own to create a more engaging and seamless
chat experience. Be sure to consider how and where you can
use data from within the Now Platform® to enhance each
conversation.

It’s also important to create a recognizable visual identity with

your company’s colors and logo while you define a personality
and tone for the Virtual Agent’s language. The tone can be
playful, professional, or somewhere in between, but it should be
applied consistently across all topics and it should always carry
empathy. Remember that users are initiating a chat because
they need help solving a problem.

Select messaging channels

Virtual Agent can be deployed to your ServiceNow Service Portal,
embedded in a web pages and mobile apps, or integrated
to where your teams already work—in products such as Slack,
Microsoft Teams, and Facebook Workplace Chat. The Virtual
Agent can be available in one or more of these channels at the
same time, so be sure to discuss which will be enabled.

5
Map conversation flows
Here you’ll create visual maps that represent the end-to-end
flow of each conversation. Virtual Agent ships with conversation
topics for IT, customer service, and HR service delivery that
you can use as out-of-the-box solutions. But you can also
use them as references or starting points for more customized
conversations. Visualizing the flow in this step will not only make
it easier to build in Virtual Agent Designer, it will also improve the
quality of the conversation by identifying any missing paths or
potential dead ends.

Considerations during the mapping step:

1. How will users initiate a conversation and which actions
will be available?

2. What responses will the Virtual Agent provide to

each question?

3. W
 here will the conversation split based on conditional logic
or a decision made by the user?

4. What Now Platform information can be used by the Virtual

Agent to personalize and improve the experience?

5. H
 ow do responses differ when a search returns no results,
one result, or more than one result?

6. W
 hen will users be automatically transferred to a live agent?

7. How will users provide feedback about their chat experience?

6
Building a Virtual Agent
Create conversation flows
Conversations are built in Virtual Agent Designer through drag and
drop functionality and JavaScript coding. The interactive canvas allows
designers to easily place prompts or responses, rearrange steps, and
preview a conversation. Business analysts and service/process owners
can build simple conversations without any code or with very little code,
while developers assist with complex flows that need scripting.

Follow configuration best practices

You can improve the quality of your conversations by following
these recommendations for Virtual Agent Designer:

• Keep your topics small and decision logic as simple as possible.

• Make conversations visually descriptive and easier to debug by

separating long blocks of JavaScript code into multiple nodes
with shorter code snippets.

• Use script variables for storing static values such as IDs, method
names, and limits instead of hardcoding values in scripts.

• Use the native text, link, card, and image outputs instead of
custom HTML to create the best visual experience across different
messaging clients.

• Avoid inline style references to CSS classes or IDs if you do have a

need to use custom HTML.

• Call any REST messages using script includes.

Test and validate

The last step before launching your Virtual Agent is to ensure it is
functioning as designed and is effectively solving problems. After
exposing the Virtual Agent to a set of test users, you should collect
feedback to see if the conversations are flowing smoothly, when
and where users might abandon a conversation, and if the users
noticed any errors. You can also make use of the gs.debug and
gs.info logging files to troubleshoot your conversations.
7
Launching the Virtual Agent
Promote adoption
Your new Virtual Agent can't support your business goals unless
customers and employees are aware it exists. Build and execute
a communication plan that leverages your ServiceNow Service
Portal, company intranet, office signage, social media, and your
web presence to increase awareness and encourage usage. With
any kind of campaign, make it easy for users to find and access the
conversations through links or QR codes that bring users directly
to the Virtual Agent. Clearly establishing a feedback loop within
or after each chat will also allow you to understand how users felt
about their experience.

Measure usage and impact

Process and service owner stakeholders can use Performance
Analytics and the built-in Virtual Agent dashboard to track the
KPIs you previously identified such as how many unique users are
having conversations, which topics are not being used, and which
messaging channels are most active. With that information,
you can make informed decisions about how to improve your
Virtual Agent.

Iterate and expand

After a successful launch, you can use the data and feedback
you’ve collected to continuously refine and improve the
conversation topics. Look for ways to reduce error messages and
minimize repeated actions, or use A/B testing to validate new
path improvements. You can also return to the design phase and
increase the value of your Virtual Agent by creating additional
topics.

8
 Creating great customer and employee
experiences with Virtual Agent
Practical guidance for a successful implementation

LEARN MORE

About ServiceNow
ServiceNow was started in 2004 with the belief that getting simple stuff done at work can be easy, and getting
complex multi-step tasks completed can be painless. From the beginning, ServiceNow envisioned a world where
anyone could create powerful workflows to get enterprise work done. Today, ServiceNow is the cloud-based
platform that simplifies the way we work. ServiceNow software automates, predicts, digitizes, and optimizes business
processes and tasks, across IT, customer service, security, human resources, and more, to create a better experience
for your employees and customers while transforming your enterprise. ServiceNow is how work gets done.

© 2019 ServiceNow, Inc. ServiceNow, the ServiceNow logo, Now, Now Platform, and other ServiceNow marks are trademarks and/or registered trademarks of ServiceNow, Inc. in the United States and/or other countries. Other company and product
names may be trademarks of the respective companies with which they are associated.

BACK SN-EB-Virtual-Agent-implementation-steps-eBook-022019