When churn is good // Throughput & graduation // KPIs for outcome-oriented products

November 10, 2022 · writing

Cancellations all look equal in your metrics, but they aren’t all the same to the member. 

For example, when a high school student “churns,” that could represent:

• Graduation (good churn)
• Dropout (bad churn), or;
• The student’s family has moved (neutral churn).

Just as churn doesn’t always signifify something bad, retention doesn’t always signify something good. when someone “retains” for another year of high school, it can be:

• Advancement to the next year (good retention)
• Flunking and repeating a year (bad retention), or;
• Necessary accommodation for a learning disability, illness, or injury (neutral retention)

While a typical SaaS subscriber is supposed to be “forever,” an outcome-oriented user is supposed to be temporary. The idea is to get them to success, as quickly as reliably as possible, and then trust that they’re going to say good things about you. (As your business evolves, you might daisy-chain multiple outcomes together to extend a member’s time with you; but you’d never want to intentionally delay a single outcome.)

For these sorts of outcome-oriented products (as opposed to engagement-oriented products), the key metrics might be:

• Graduation Rate (% of users who reach successful outcome)
• Throughput Speed (how quickly they can do so)
• Referral Rate (expected # of referrals per successful graduate)
• Repetition Rate (harder to measure and not applicable to every outcome-oriented offering, but represents the % chance that, should they reencounter the problem, they’ll return to use your thing again)

If the above is broadly accurate, it would suggest that the growth engine of outcome-oriented products behaves less like the long slow SaaS ramp of death and more like a modified viral loop. (As a reminder, the viral loop is basically the expected # of referrals per user, multiplied by your conversion rate of a qualified lead, accelerated or delayed by the time until referral.)

The delay before referral is important and often overlooked. One of the reasons social games like Farmville could grow so fast was that the referral happened within the first hour or so.

In the case of offering an inherently lengthy process (like a long-form course or community), the time until referral is already going to be quite slow. This means you wouldn’t want to try to “extend retention” by delaying value in any way. If you can front-load the value, if you can give it to them faster, then even if they do unsubscribe (as successful graduates, i.e., good churn), that faster “loss” in subscriber will be immediately compensating by the faster gain of an extra referrer. 

So I think it’s possible, in this category of businesses/product, to find yourself in a situation where “lower” retention (via faster throughput and high graduation rate) equals higher growth.

← all writing