How Advanced Service Desk Chatbots Work
by Markos Symeonides, on 10-Sep-2020 15:45:00
A service desk chatbot feeds on data. The more data you feed it, the more it can do—increasing value for customers and reducing costs for IT.
Seamless integration into your knowledge base, infrastructure, service and support data unlocks a chatbot's ability to handle a broader range of incidents, requests, and inquiries.
Part of our Omnichannel ITSM series.
In the last article How to Launch a Service Desk Chatbot we talked about how an agile approach is the key to success with chatbots: Start with a Minimum Viable Product (MVP) chatbot that delivers some early value to customer and then iterate to add advanced capabilities.
Customer expectations grow fast, so you'll need to ramp up your service desk chatbot capabilities quickly. This is how you gain traction and make a chatbot the channel of choice for IT customers. The more it can do for them, the more confident they will be that the chatbot is the right channel, and the more frequently they will choose it.
However, the speed at which you can add new chatbot capabilities depends on how easily you can connect your chatbot with the information it needs. The number of issues and requests a chatbot can handle without human intervention depends on how much information you give it to work with, and whether you give it actionable control.
Chatbots and Incident Management
Many organizations start by deploying a chatbot which connects to their support knowledge base because it's a ready source of structured support data that can be returned as "how to" guides. This gives customers a novel chatbot experience—but really it's just adding a natural language search engine to the front-end of your knowledge base. The main advantage here is that they don't need to click around so much to get the information they need. The experience is a little newer and a little faster.
But this is hardly a game-changer. The game-changing chatbot capabilities come from connecting it deep into your ITSM data and empowering your chatbot to act to solve customer issues and deliver services in a completely digital, automated way.
Depending on the size and health of your knowledge base, using it to fuel your chatbot might help it solve 10% of your customers' support issues—but that won't be enough to drive mainstream adoption. To achieve mainstream adoption, you'll need to extend capabilities so that users are confident that the service desk chatbot will be able to help them in most cases—otherwise they won't even try for fear of wasting time.
NOTE: From the very start you need to make it easy for your customers to self-escalate an issue to a human agent so that if they can’t get what they need from your chatbot, they have a frictionless path to speaking with a human being.
Chatbots and Request Management
If your service desk chatbot doesn't connect to your service catalog, it won't be able to match customer requests with available services and it will be stuck in "Incident mode": It will be able to help customers having issues with the services they already use, but it won't be able to help them get what they need—constraining the value that your chatbot can deliver.
Integration between a chatbot and a service catalog takes value to the next level by providing a new and different experience for customers: they can use natural language to ask for what they need—and the chatbot will use Natural Language Understanding (NLU) and fuzzy logic to put the solution right in front of them.
For the service desk, it means large chunk of "I need..." service requests can be diverted from the expensive phone channel to the more economic chatbot channel.
There are a couple of levels that can be applied here, allowing for an agile approach:
- The chatbot analyses the request, searches the service catalog for a suitable services, and shares a link with the customer so they can go right to the page where they request the service.
- The chatbot is connected into the ITSM platform workflow system, empowering it to trigger the service delivery workflow directly—saving the customer a step.
Of course, alongside the data that describes the service, you will need decent service descriptors so that your bot can effectively match user needs with service solutions. Most organizations need to improve the way they describe services in the service catalog—for both people and machines.
Chatbots and Inquiry Management
Service desk's tend to focus on solving technical issues, but information requests can be a large chunk of the total volume of calls. Inquiries can have a broad range, from "What's the status of this service?" to "What's happening to incident #234869?".
To find answers to different types of inquiries, your chatbot will need to be able to reach into various ITSM datasets. For example:
- Answering a service status question means the chatbot must have visibility of service performance and infrastructure monitoring data.
- Answering incident and request status queries requires visibility of support records—specifically the trail of activity so that the bot can report back to the user.
Advanced inquiry handling capabilities take the full context into consideration to make it an even better experience. For example, simply asking "What's the status on my incident?" will trigger the chatbot to check incidents logged against that user and then either give an answer if there is only one, or let them choose which one they are referring to if there are more than one. This means the user doesn't need to search for their case number. Context-aware service desk chatbots eliminate friction from the IT customer experience.
Advanced Chatbots Rely on Unified ITSM
A service desk chatbot can do so much more for your IT customers when it's powered by a full set of rich ITSM data. But for many organizations with legacy ITSM solutions, it's difficult to make those connections—multiple modules duplicate integrations efforts and create a complex chatbot architecture which is difficult to maintain. Some capabilities may even break when modules are upgraded.
The answer is the simplicity of a unified ITSM platform.
Our assyst ITSM platform is different. It's a single app with a single central CMDB which holds all ITSM data in one data model. That means there is one connection between your chatbot and all the data it needs to unlock the advanced chatbot capabilities that can make a real difference to your service desk and your business.
And even though it's AI, it doesn't require a team of egg-headed boffins in lab coats to make it work. A full chatbot deployment can be done in just six weeks—diverting 25%+ of your service desk calls to the streamlined chatbot channel.
Contact us for a no-obligation chat with a consultant
Sign-up for blog notifications to make sure you don't miss out
More About Omnichannel ITSM
- Why You Need an Omnichannel Service Desk
- How Omnichannel ITSM Improves Quality, Speed, and Cost
- Multichannel vs Omnichannel Support: What's the Difference?
- Omnichannel Service Desk Strategy: 3 Things You Need to Know
- Multichannel / Omnichannel Service Desk Adoption
- How to Make Omnichannel ITSM Work
- How to Launch a Service Desk chatbot