I finally found the product metric I’ve been looking for

As a product leader, I’ve spent years elusively chasing the perfect product metric. Something that was high-level enough to give a good idea of how the overall product is performing, but also granular enough to see what levers need to be pulled to make improvement. Something that was a leading indicator to churn and to future growth. Something that didn’t take hours of analytics, BI jujitsu and double PhDs from MIT to calculate. And I “think” I’ve found it…

Two disclaimers. 1) No metric is actually perfect. 2) I say I “think” I’ve found it because we just started using this metric and time will tell if it’s as effective as I think it will be.

Let me introduce you to the Product Engagement Score or PES. PES is a composite score that combines adoption, stickiness, and growth into a score on a scale of 0 to 100. I’ll dive into each component of the score, but the reason I love the overall concept is that it is the cleanest metric I’ve seen that combines adoption and repeat user behavior in a way that is a leading indicator to churn and user growth. Full disclosure, this metric was created by Pendo, the product analytics SaaS company. We are customer’s of Pendo, but this article was written independent of any influence from their company.

Here are the components of PES:

Adoption

Adoption is a roll up of the most critical features in your product. It takes the number of accounts (or users — depending how you want to set it up) that have used that feature divided by total accounts (or users). Then you average the percentages together for the adoption score, which is a percentage from 0 to 100%. The hardest part about setting up the adoption score is determining which features to include and making sure you have the proper tracking of these features. Best practice is to not use more than 10 core features.

Here’s a peek into our process for deciding our core features. We listed all the features in our product and then each product manager ranked their top 8 features. We had each PM do their ranking in a separate spreadsheet to avoid groupthink. Then we compiled the scores and debated across the team. Some features were removed because they only applied to a segment of our user base and not the entire base. Other features were important, but don’t have great tracking (the maturity of this process is to get better tracking and then include those features). We ended up with seven core features with good, trackable data. One note is that for some features we had to find all the spots in the product they could use the feature and create a roll up “master feature” for tracking purposes. For example, a user can create an invoice from four different places in the app, and we want to track all those scenarios. Our guiding principle was to ask this question, “If a customer is using these features does that mean they are a healthy account getting value from our product?”

Stickiness

Stickiness is also a percentage from 0 to 100% and measures daily active accounts (or users) over monthly active accounts (or users). You’re trying to see how frequently your active users are in the product. An account that logs in 18 business days a month versus 3 business days a month is considered stickier and healthier. You can adjust the ratio to whatever makes sense for your business. You can use daily over weekly or weekly over monthly.

Growth

The growth metric is a ratio of new accounts plus recovered accounts (or users) over dropped accounts (or users). Here are definitions of each metric:

• New — first event is in the time period
• Recovered — user or account didn’t have activity last time period but does in the current time period
• Dropped — user or account that was active last time period but is not active in the current time period

Seasonality could affect this number if users grow or shrink or if accounts cancel and then come back. This metric should correlate to your Net Recurring Revenue number although the PES growth metric is not revenue based.

My biggest disappointment so far is that because this metric is new, there aren’t great benchmarks. But in reality, because every business is so different and there are different ways to customize the PES calculation, benchmarks may not mean much anyway. The perspective we’re taking is a focus on improving the number. We set our baseline at the current PES score and we’re trying to improve it by 10%. And each of the three components of PES become the levers we can pull to impact the overall score.

I would encourage you to dig into PES and see if it might just be the product metric you’ve also been looking for.