Measuring feature usage and user adoption in SAP S...

BenMcGuire · ‎03-17-2023

Introduction

User adoption is a hot topic where I am working. We are following the usual user adoption metrics but we wanted to take it one step further. The data source Response Time Analysis [COD_RESP_TIME_ANA] is only containing the last 30 days data. If you have a high number of usage or a large number of interactions (>250k records) it can be challenging to do any meaningful analysis on usage of the system.

Understanding not just what objects users are creating (opportunities, Leads, Visits) but HOW they are using the system can give us some great insight into areas of improvement or help us prioritise where we put our efforts.

Objective of this blog

Show how to Expose this data source to an external BI Solution (example: Snowflake).
Show examples of reports built in Power BI.

Expose the Data source to your external BI Solution

Navigate to Business Analytics > Design Data Sources. Select the data source COD_RESP_TIME_ANA and click on expose. Now you can start to consume the data source in your chosen BI solution. In the examples shown we are taking a daily snapshot of data for the last six months. Since we are using a data lake you can also enhance the data by joining with other external HR information (Country, role type etc) or other C4C data (business role, Sales Org etc.)

Expose Data Source in C4C

Examples of reports built in Power BI

With this example you can get details of a system usage over time. In the example we look at 'Work Center' displayed over the day. Where our peak system usage is and in which objects in C4C whilst also summarizing the most used Interaction steps and which clients are used.

With this example you can get details of a feature usage over time. In the example we look at 'Export to Microsoft Excel'. Since this tracks the user ID we can also include the business role of the user executing the interaction step.

In the below example we look at 'OVSValueSelect' interaction step. Users selecting values in a click box. We can see large usage in the Service Ticket area where users are categorising tickets. Could this be something to be improved? Since we are using machine learning for ticket categorisation could this indicate the model needs to be retrained?

Summary

Analysing this data raises a lot of interesting discussions with users, stakeholders and development teams. Also check out this other user adoption metric blog. Something to consider is security and who is able to analyse this type of data. Check in your local region if you are able to track this type of usage data or consider anonymising the user data.

I encourage you to share your thoughts on this blog, share the metrics, KPIs and reports that you already track or want to track within your organization, how the metrics help you in driving adoption and what more you’d like to see as additional content in later blogs

Ben

Saurabh_Kabra · ‎03-17-2023

Hi Ben,

Thanks for the summary. I have a few questions though as I was recently working ona similar topic of user adoption... How do you get the data from C4C in Snowflake or powerBI using odata?(i mean, Do you use the live connectivity or you store them in thier local DB via periodic job?)

Also in my recent project, we tried to use OData periodically but they were running into errors due to the sheer amount of data. Did you also face such issues and how did you overcome them?

Best

Saurabh

BenMcGuire · ‎03-20-2023

Hello Saurabh.

We get the data from C4C using standard integration to SAP BW. From there the data is sent to snowflake/MS Azure and then consumed as a data set in Power Bi. All the data joins are done in snowflake. Its a periodic job that ingests the data into SAP BW.

As you said the dataset can be quite large. For 3500 active users and 6 months history the data source is around 15m+ records.

Saurabh_Kabra · ‎03-20-2023

intersting!

VinitaSinha · ‎03-29-2023

Love it Ben ! Especially the OVS analysis and using these insights to evaluate the possibility to re-train the ML model.

Also love that you took the example where 30 days data can be exported out and built over a period of time to do analysis in a BI environment

Thanks for sharing these insights !

BenMcGuire · ‎03-30-2023

Thanks vinita.sinha !

amit_dudhbade · ‎09-29-2023

Hi Ben,

Your blog post is not reaching all the CX Community members. Please use the SAP Managed tags to make your post visible on CX Community Pages. Check this blog post for details https://blogs.sap.com/2023/09/29/how-to-make-your-blog-posts-visible-on-sap-customer-experience-comm...

Hope this helps.

Best regards,

Amit