A practical guide to ServiceNow chatbot configuration (2025)

Table of Contents

• What is the ServiceNow Virtual Agent?
• The moving parts of a ServiceNow chatbot configuration
  • Initial setup: Waking up the Virtual Agent and NLU
  • Building conversations in the Virtual Agent Designer
  • Connecting channels and tools
• A ServiceNow chatbot is only as smart as its knowledge
  • Using the native ServiceNow Knowledge Management base
  • The challenge of outside knowledge
  • A unified approach can be simpler and better
• Advanced features and pricing
  • Using generative AI with Now Assist
  • Making sense of the ServiceNow CSM pricing model
  • Hidden costs and other options
• Is there a simpler, more powerful way?
• Get started with AI that actually works for you

ServiceNow is a beast of a platform, and for good reason. It’s the backbone for countless enterprise workflows, from IT service management (ITSM) to customer support. Its native chatbot, the Virtual Agent, promises to automate frontline support and free up your teams. But getting it up and running isn't always as simple as flipping a switch. The configuration process has some important details you'll want to know about before you dive in.

This guide will give you a clear, no-nonsense overview of ServiceNow chatbot configuration. We'll walk through the initial setup, how the bot connects to your knowledge, its more advanced AI features, and the often-confusing pricing tiers. We'll also look at some simpler, more flexible alternatives that might get you to your automation goals a whole lot faster.

What is the ServiceNow Virtual Agent?

The ServiceNow Virtual Agent is the platform's built-in chatbot, designed to handle common requests, answer user questions, and automate simple tasks right inside the ServiceNow world. Think of it as the digital first line of defense for your support channels, whether they're for internal employees needing IT help or external customers looking for answers. Its main job is to offer 24/7 self-service.

At its core, the Virtual Agent works by following conversational workflows called "topics." For instance, a "Password Reset" topic can guide a user through the steps to reset their password without ever needing to talk to a person. The bot uses Natural Language Understanding (NLU) to figure out what someone is asking for, but getting its full brainpower often means upgrading to a pricier plan.

Because it’s baked into the platform, it connects nicely with other ServiceNow modules like the Knowledge Management base, Service Catalog, and Incident Management. This lets it do things like create a support ticket or find a help article, as long as all that information and those workflows are already set up inside ServiceNow.

The moving parts of a ServiceNow chatbot configuration

Getting a ServiceNow chatbot set up involves a few key steps and pieces. While the platform gives you some powerful tools, it's good to know that each layer can add a bit of complexity to the whole project.

Initial setup: Waking up the Virtual Agent and NLU

Your journey with a ServiceNow chatbot configuration starts in the plugin menu. You'll need an admin to find and activate the right plugins, like "Glide Virtual Agent", just to turn the chatbot on. But right away, you hit your first big decision: Virtual Agent Lite vs. Pro.

If you’re on a standard ServiceNow license (like ITSM Standard), you get Virtual Agent Lite. The "Lite" version comes with some pretty big guardrails. The main one is that you can't create your own custom conversation topics from scratch; you can only copy and tweak the handful that come included. This means any unique process you want to automate is off-limits unless you upgrade. Full NLU capabilities are also turned off, walling off true customization from day one.

For those with a Pro license, you get access to the NLU Workbench, which is the tool for training your chatbot to understand different user requests. This is a must for building an intelligent bot, but it has its own setup process and requires a real commitment to ongoing tuning to keep it accurate.

Building conversations in the Virtual Agent Designer

Once it’s activated, you’ll spend most of your time in the Virtual Agent Designer. This is a low-code, drag-and-drop tool for building and editing the conversational flows, or "topics," that your chatbot follows.

ServiceNow gives you a head start with a library of pre-built conversation templates for common ITSM, HR, and CSM tasks. This is genuinely helpful for handling standard stuff like checking a ticket's status or searching the knowledge base.

However, that "low-code" promise can get a bit blurry. The moment you need to automate a process that isn't in a template or involves tricky logic, you'll probably find yourself needing to write scripts. Creating truly custom and smart workflows often demands a deep understanding of the ServiceNow platform and some developer help, which can push it beyond the reach of non-technical teams.

Connecting channels and tools

You can put the Virtual Agent in different places, including the native Service Portal, the ServiceNow mobile apps, and popular workplace tools like Slack and Microsoft Teams. This is great for meeting people where they already work.

The tricky part comes when you need to connect to tools and data sources outside the ServiceNow bubble. To do that, you usually have to use the ServiceNow Integration Hub, which is a powerful but separate platform for building connectors (they call them "spokes"). If a pre-built spoke doesn't exist for a tool you use, you’ll need a developer to build one, which can add a lot of time, complexity, and cost to your project.

Some newer tools, like eesel AI, go a different route by offering one-click integrations that connect directly to your existing helpdesk and knowledge sources, without making you use a separate, complicated integration platform.

A ServiceNow chatbot is only as smart as its knowledge

A chatbot can't answer questions if it doesn't have access to the answers. For a ServiceNow chatbot configuration, this means you have to think hard about where your knowledge lives and how the bot can get to it.

Using the native ServiceNow Knowledge Management base

Right out of the box, the Virtual Agent is built to work with the ServiceNow Knowledge Management (KM) module. This is its main source of truth. When a user asks a question, the bot's first move is to search the KM base for relevant articles to summarize or share.

This makes having a clean, complete, and up-to-date knowledge base in ServiceNow an absolute must. If your team's collective wisdom isn't already documented inside the ServiceNow KM, your chatbot will be flying blind from the start, and it won't be very helpful.

The challenge of outside knowledge

Here’s a huge hurdle for most companies: knowledge is almost never in just one place. It's scattered across platforms like Confluence, Google Docs, Notion, SharePoint, and, most importantly, buried in the resolutions of thousands of past support tickets.

Connecting these outside sources to the ServiceNow Virtual Agent is not a simple, built-in feature. It’s a big project that often requires custom development, expensive third-party connectors, or painful, manual data migration. This creates a knowledge silo where the chatbot can only see a tiny fraction of your company's actual knowledge. The result? The dreaded "I'm sorry, I can't help with that," which just sends users right back to your human agents.

A unified approach can be simpler and better

This is where a different style of AI can make all the difference. An AI Agent from eesel AI is designed to solve this exact problem by bringing all your knowledge together instantly.

With one click, it connects not only to your ServiceNow knowledge base but also to all the other places your information lives, Confluence, Google Docs, past tickets from helpdesks like Zendesk, and over 100 other sources. Instead of forcing you to move all your knowledge into one system, modern AI platforms can securely access information wherever it is. This gives the AI a complete picture so it can provide accurate, helpful answers from day one.

Advanced features and pricing

Getting a basic, keyword-driven chatbot running in ServiceNow is one thing. But unlocking its true potential with generative AI means looking at more advanced features and, of course, a more complicated pricing structure.

Using generative AI with Now Assist

For a truly conversational experience powered by generative AI, you need to look beyond the standard Virtual Agent. ServiceNow's premium solution is called Now Assist for Virtual Agent.

This is a separate, high-end product that uses Large Language Models (LLMs), like the ones from Azure OpenAI, to give dynamic, summarized answers instead of just links to articles. It can understand more nuanced conversations and generate human-like responses. But this powerful capability is an add-on that requires a CSM Pro Plus or Enterprise Plus license, putting it at the very top of ServiceNow’s pricing ladder.

Making sense of the ServiceNow CSM pricing model

ServiceNow’s pricing can be notoriously complex, with features split across different packages. To get the chatbot you actually want, you need to pay close attention to which plan you're on.

Here’s a simplified breakdown based on their Customer Service Management (CSM) packages:

Source: Based on data from ServiceNow's Customer Service Management pricing page.

So what does this really mean? To get a chatbot that you can fully customize (the full Virtual Agent), you need to be on at least the Professional package. To get the modern, generative AI-powered experience that users expect these days, you need to buy an additional Plus add-on. This tiered model means the total cost can climb pretty quickly.

Hidden costs and other options

Beyond the license fees, there are other costs to think about with a ServiceNow chatbot configuration. These include the specialized admin or developer time needed for setup and ongoing maintenance, and the long wait to see value as you build and tune conversations. There's also the risk of paying for a lower tier only to find out its limits force you into an expensive upgrade later.

This is quite different from the straightforward pricing of a platform like eesel AI. Our plans are based on the features and capacity you need, with no surprise per-resolution fees. You won’t get a shock at the end of a busy month, which lets you scale your support automation without worrying about the bill.

Is there a simpler, more powerful way?

While ServiceNow is an incredibly robust platform, its native chatbot configuration can be a slow, complex, and expensive journey. For teams that want a faster path to seeing a return on their investment, a modern AI agent that layers on top of your existing tools, including ServiceNow, can be a great alternative.

The eesel AI Agent is built from the ground up to be simple, flexible, and powerful.

• Go live in minutes, not months. Forget about long implementation projects. eesel AI offers a self-serve setup with one-click integrations, so you can be up and running in the time it takes to finish your coffee.

• Unify all your knowledge instantly. Don't worry about where your information is. eesel AI connects to all your knowledge sources from day one, including ServiceNow, Confluence, and past tickets, avoiding the data silo problem completely.

• Test everything with confidence. Worried about letting an AI talk to your users? eesel AI’s simulation mode lets you test your agent on thousands of historical tickets in a safe environment, so you can see exactly how it will perform before you go live.

Get started with AI that actually works for you

A successful ServiceNow chatbot configuration is definitely possible, but it demands a real investment in licensing, setup time, and getting all your knowledge in one place. It's a solid tool for organizations that are deeply committed to the ServiceNow ecosystem and have the resources to support it.

However, for teams that need a faster, more flexible, and more affordable solution, modern AI platforms that integrate with your existing tools provide a much clearer path forward. By unifying your knowledge and working with your current setup, you can start automating support and making users happy right away.

Ready to try an AI agent that connects all your knowledge and starts resolving tickets in minutes? Try eesel AI for free or book a demo to see how it works.