Q for IHers/PMs running SaaS products with dozens/100s/1000s of customers
How do you measure adoption (use) of all the features you build?
New feature development is a constant thing in SaaS. Measuring adoption is important when you're pushing out so many new features -- is what we're building actually being used as much as we expect it to?
We instrumented and built some internal dashboards to track feature adoption (usage) across accounts and users. It was a homegrown thing though and mostly tracked # of users having done something in a certain period of time - eg leads created, invoices sent etc. I want to move away from the homegrown to using a tool.
What are tools or softwares you use to do this? Does your product team meticulously track feature adoption and usage post-rollout and how?
We started to include features for tracking engagement (https://djaodjin.com/docs/djaoapp-api/#proxy_engagement_users_list). We do that by adding engagement tags to URL paths and measuring engagement very early on during HTTP request processing (before business logic).
It also has the side-effect of being able to trigger onboarding tutorials first time a user engages with a feature/URL path pattern.
This is interesting, I've always had a weird view on features in the following sense. If we build at feature is needs to move profit or revenue. This normally forced me to create reports based on some event being created, e.g. someone created a text campaign and sent it. How do you feel these engagement tags have helped make your reporting easier? Do you use system logging tools to count the times the user calls the endpoint?
First with current traction and our small team, we are doing pain-driven development. Basically for the past months, we haven't had time to work on anything else that what customers complain the loudest about. As a result, a significant amount of time is spent in DevOps work, keeping the site responsive under increasing load, which does not translate well into "measuring features used".
Engagement tags are indeed measured through real-time log analysis. They have been useful to understand how specific users moves through the onboarding pipeline. That gives us insight on when to talk to them, which tutorials to improve, and what features to advertise.