on 09-08-2019 07:18 AM - edited on 03-06-2023 07:29 PM by Victor Chen
How to decide what to build first for your Virtual Agent?
You likely have the answer right in front of you – and it’s in your data.
It all starts with mining relevant data and aligning that data to your goal or set of requirements. Are you looking to drive down volume of common service requests, deliver a great experience to employees, align service delivery across departments, or all of the above? These are the types of questions to explore further to ensure you have the right metrics to drive expected outcomes.
Being data-driven in defining and prioritizing your Virtual Agent topics to create strategic, automated experiences for your users is a repeatable strategy for success. Focusing on the top 20% of use cases that are happening 80% of the time will reduce the time to achieve your target, and provide the foundation necessary for overall adoption.
Let’s expand on the 80/20 rule further. By analyzing the top 20% of historical incidents, cases, requests, and live agent chat message exchanges, your data should provide you with a clear delineation of the top use cases that take up a disproportionate amount of your agents’ time and productivity. Password resets, VPN issues, status checks on a ticket or case, and help with pay-related matters are just some of the most common use cases customers tend to report.
How Virtual Agent can help find your top 20%
It turns out you are actually sitting on a goldmine of data that can form the way employees interact with service delivery and their organization in its entirety.
To help jumpstart your Virtual Agent journey, ServiceNow provides many out-of-the-box topics of frequently requested, high volume services.
Topic Recommendations - identify your top Incident drivers
Topic Recommendations uses your organization's historical data to find the most relevant pre-built topics that your organization can benefit from, and new topics that would be useful to create. You can analyze Incidents, Requests, live chat transcripts, and user utterances. After you run the Topic Recommendations analysis, it will display recommended pre-built topics to add to your Virtual Agent plus the NLU intent. For more information, see the documentation on Topic Recommendations here.
Requirements to use Topic Recommendations:
See the Virtual Agent Academy on Topic Recommendations, including configuration steps and demo.
Using Performance Analytics to prioritize topics
Aside from Topic Recommendations, you can also build analytics reports on Incident or Case data to find issue hotspots for the Virtual Agent to address. See the below Academy video for examples.
Conversational Analytics Dashboard
With continual improvement and excellent employee experiences being top of mind, use the Conversational Analytics Dashboard, also available as of Quebec, to collect usage trends and identify blue sky areas for optimized service and automation.
In summary, first, determine what to solve for, and then start small by simply identifying 5 use cases i.e. that will yield your organization the highest value. Think MVP (minimum viable product). Ideal candidates include specific services, commonly submitted incidents, and any identifiable trends critical to your users' productivity. Once these 5 use cases are identified, implement, test, and repeat the process to enhance the impact and adoption of your Virtual Agent experience.