16. Using State to Retain Customers 🧱
Two types of retention plus one strategy and four examples in the real world.
Hi friends 👋
It’s been awhile! I hope your 2023 is off to a great start. I’m (finally) making time to write more. Back on the every Sunday train!
Recently, I moved into a Product role at Ness. Topics are going to shift a bit here as I learn what it means to be a great product leader. As always, really appreciate you following along.
Today’s read is 4 minutes long and covers two types of retention plus a strategy and some examples.
See you next Sunday! Feedback encouraged.
Jeremey
Two types of retention
Stacked against the other pirate metrics, retention both tougher to measure and not quite as fun to work on compared with, say, acquisition. As a result, there aren’t bountiful retention playbooks, frameworks, or easy-to-describe strategies.
Today, we’re going to talk about state, which is one such framework. But first, let’s talk about the different types of retention.
Patrick Campbell (CEO of ProfitWell) recently broke it down into two areas—strategic versus tactical retention.
Strategic - Driving more value for customers through what you build (e.g. your product roadmap).
Tactical - Addressing “mundane” opportunities—payment failures, broken upgrade frameworks, bugs, etc—that cause users to churn.
Today’s topic is focused on the strategic side, but this distinction gave me a new appreciation for tactical retention. We’ll jam on that later.
State as a strategic retention tactic
I was introduced to the concept of state from Julian Shapiro.
Think of state as non-transferable status of some kind that makes your product incredibly sticky.
The concept was inspired by video games. As you play, you accrue state in the form of extra powers, rewards, special weapons, etc. You can’t take these extras with you if you move to another game.
As Julian notes, state works for two reasons:
Players don’t want to lose everything they’ve worked for. Loss aversion is a real thing.
As you accumulate state, you can exploit it to accrue more. The rich keep getting richer.
We can apply the concept of state to products.
State in the real world
Let’s walk through some real world examples of how this comes to life.
Reputation (Uber, Facebook Marketplace, Etsy)
Drivers spend years building up a record of 5-star reviews. This pristine record then unlocks more opportunities as better drivers get more rides. You can’t export these reviews to any other platform (e.g. Lyft).
Captive audience (Substack, YouTube, Twitter)
You can’t move your Twitter, Facebook, YouTube, etc audience elsewhere. That’s obvious.
I want to focus on Substack because they’ve introduced state in a genius way with Recommendations.
Two things happen when a writer recommends a publication on Substack.
Their existing audience is notified of the recommended publication (drives subscriptions).
The writer of the recommended publication gets an email notification (builds community and goodwill).
You can export your email list, but you can’t export this ecosystem of recommendations.
Unique infrastructure (Zapier, AWS, Stripe)
Anyone can connect two APIs. Heck, even three or four. It’s much more challenging to connect 5,000+ applications all on an infrastructure that supports millions of concurrent connections.
When you connect your applications through Zapier, two things happen:
You build a series of zaps that automate your work. They’re familiar enough for anyone to use and proprietary enough that it’s tricky to move them elsewhere.
Your zaps just work in a reliable, predictable way.
This second piece is often overlooked, but it creates a moat for Zapier. This kind of technical infrastructure is very difficult to build. Once these systems are embedded in your workflows, it’s very hard to pull them out.
Historical data (Strava, YNAB, MyFitnessPal)
Millions of endurance athletes around the world track the minutiae of their daily workouts on Strava. They turn all of this data into historical insights and performance metrics.
Want to rank your last run against last month? Simple. Track your time on a specific course over years? Easy.
Of course, if you leave Strava, it’s very difficult (impossible?) to take this analysis with you since it’s built brick-by-brick over time.
The beauty is in the combination
State becomes more powerful as you combine strategies. Back to the Strava example, they use several forms:
Historical data in the form of training analysis.
Followers and kudos give you a non-transferrable audience that applauds each of your activities.
Strava segments introduce a series of achievements (King/Queen of a specific route). This feeds into a reputation that exists only on Strava.
It’s not trivial to connect to all of your various devices (watch, bike computer, etc) and ingest data easily—a play on the infrastructure piece above. Strava, in a sense, is the foundation for your activities.
These all play together within the product to retain users for the long haul. When done well (as in the case of Strava), this all feels natural and works for both the customer and the business.
What else?
Flat backlogs are tough to prioritize. Here’s a useful analogy comparing user stories to a skeleton, which gives them shape and helps you prioritize and plan releases.
Are we stat sig yet? Probably not, and it might not matter with experimentation.
I’ve been interviewing a lot of users lately and really digging this approach for prioritizing opportunities, not solutions from Teresa Torres.
WTF is strategy? Specifically loved the Wheel of Fortune reference.