Automation Discovery uses machine learning to identify potential automation opportunities in your HR case data. For example, Automation Discovery can help you identify potential Virtual Agent conversations and workflow opportunities that you can use to deflect future cases and reduce case load for your agents. The 2023 store app release provides a robust HR taxonomy to identify automation opportunities in your HR case data. Note – Automation Discovery was released in the Spring of 2021 with an ITSM taxonomy used to categorized incidents.
To install Automation Discovery
1. Ask your admin to go to the ServiceNow store.
2. Once on the store search for “Automation Discovery”
3. Before installing Automation Discovery also install the dependencies. You will need Virtual Agent and the NLU workbench. You will also need to install the HR Service Delivery NLU Model for Virtual Agent Conversations and Employee Experience VA topics and topic blocks plugins.
If you don’t own Virtual Agent you can install Automation Discovery into sub production.
Building you first Automation Discovery Report for HR
Automation Discovery was designed to mimic the report building process, so if you know how to create a report in ServiceNow you can create an Automation Discovery report.
1. Go to All > Automation Discovery > Discovery Reports
2. Click New Report
3. Using the below image as a guide, name your report then (1) select HRSD as the taxonomy. This will group your data under a HR automation opportunity category. (2) Make sure the HR case table is selected. (3) Set the filter conditions, you will need at least 10k rows and no more than 100k rows – these are the defaults. (4) Select the text field to analyze, typically short description/description.
4. At the bottom click Run Report. This will take about 20minutes to complete.
Interpret your First Automation Discovery Report
1. Go to All > Automation Discovery > Discovery Reports
2. Click on your HR Automation Discovery Report
3. Reference the image below. The Automation Discovery report opens and displays three Key Performance Indicators at the top. (1) represents the HR automation opportunities in your data. In my screen shot below I have 37 HR automation opportunities with 21 that can be potentially deflected using pre-built VA topics. (2) I have 1,784 possible deflections based on those cases that can potentially be deflected by VA. Finally, in (3) this shows the time savings if you act on the top 10 automation opportunities.
4. You can flip any of the Key Performance indicators to get additional detail.
5. Scroll further down to the Automation Opportunities tab, this is a ranked list of your cases which match the HR automation opportunities taxonomy. This taxonomy is managed by ServiceNow and cannot be customized. If there is a checkmark in the Virtual Agent ready column it means, we have an OOTB Virtual Agent conversation that may be able to automate & deflect the issue. If there isn’t a check mark you can decide if it makes sense to make a custom Virtual Conversation or workflow to automate these cases. In my image below the first automation opportunity identified is HR-Talent Management-Onboarding-Onboarding Tasks, and the check mark of Virtual Agent ready means that there is a Virtual Agent conversation that may be able to automate these tasks.
6. By clicking on the carrot next to HR-Talent Management-Onboarding-Onboarding Tasks I can expand the first row to get additional detail. Reading from left to right; I can see that for this group of cases the MTTR is 6days plus, we can see the assignment group distribution, and the short description of the cases. We can click the View More Details link to get into a list view of the cases and apply filters to better understand the composition of the cases. By looking at this data we can decide if it makes sense to automate. If it does we can take the action to either Flag or Open Topic Recommendation which will take us into the Virtual Agent designer. Or we can Hide this Row if it doesn’t make sense to automate this finding.
7. The Not-Categorized tab contains HR cases which we were not able to assign to the HR taxonomy, instead we used clustering to group the records. You can use this cluster to identify if it makes sense to automate these cases.
8. At the top of the Automation Discovery report you can share your report or add watchers to review the analysis.
I’m excited that the new HR taxonomy in Automation Discovery will help our customers quickly identify where to focus their automation efforts. I recommend running Automation Discovery throughout your Virtual Agent design process so you can use the data to determine where to prioritize your automation efforts. If you don’t own Virtual Agent or Predictive Intelligence you can use Automation Discovery to help build a business case. Automation Discovery provides the case volume and mean time to resolve which are key variables used to determine how much time and money you can save through automation. -Lener
