5 of my favorite new retention techniques that you may not have tried
Hot off the press and recently tested.
Retention is life. It’s the ultimate Product-Market Fit feedback loop. It’s how you know whether your product is actually lovable or not. This means the opportunities for improvement are endless. And even though I’ve been obsessed with improving paid user retention for a long long long long time, I’m still finding new approaches.
Here are a few that we’ve recently tested… which all yielded a meaningful lift!
Plus, 2 laws of retention improvement that everyone keeps forgetting.
Sponsored by Churnkey, the platform that helps companies save 20–40% of subscription revenue that would otherwise be lost to churn. Used by best-in-class companies like Superhuman to proactively improve retention. Their new Feedback AI even maps churn feedback to MRR lost so you know exactly what to build next. Learn more →
#1 - Adding a ‘lite’ plan to the cancellation flow
When you’re doing a cancellation flow, the most important thing to ask at minimum is: ‘Why are you leaving?’ It’s so important for you to understand what’s driving this decision for the user. As I’ve said before:
Always, always ask why people are churning. Most reasons fall into these buckets and have the following distribution of responses:
“Too expensive” and “Don’t need it anymore” will almost always be your top reasons for churn. This is where discounting and engagement efforts should deliver a lot of value. The rest are red flags for your core product team to deal with.
If you’re trying to save a customer who’s no longer an active day-to-day user, a lever to pull is introducing what I call the Keep the Lights On (KTLO!) plan. This is a plan for customers who still have things running on the platform or need minimal ongoing access, but don’t need full usage anymore. They keep the core functionality they rely on, while shedding the heavier, more expensive usage layer.
So at Lovable, instead of saying ‘Sorry, you either pay the full $25/month price or lose all of the functionality’... we created a Lite Plan that allows users to keep access to the paid features only, but without any additional credits. It’s only $5/month and is offered within the downgrade flow.
When we tested this, we tried $15, $10, and $5 price points. We landed on $5 because it optimized for take rate (~10%) and user retention. Short-term revenue was lower with lower price points, but once renewals were factored in, it was net positive.
This ends up feeling like a win for everyone: Don’t need the main meat of what the paid plan offers, but want some of the paid features to keep your projects alive? Just use the Lite plan.
Note that we don’t show the Lite plan on the main pricing page. It’s not a strong free-to-paid conversion plan, because we want paid subscribers to experience the full power of Lovable, and that requires credits that come with our initial Pro offering. Instead, it’s a much better fit for someone who has already done their main building and just wants to keep the lights on.
Pro Tip: Want to try this out for yourself? This week’s sponsor, Churnkey, has a ‘Hidden Plan’ capability that makes this super easy. Check it out, here.
Although I’ve certainly seen it before, this is the first time I’ve tried it and consider me impressed.
#2 - In-app prompts for involuntary churn (payment issues)
Okay, this one isn’t a big surprise. It’s an oldie but a goodie, but we actually just tested it at Lovable. And it’s mind-boggling how much of a difference it makes.
So, in every single product, in every single subscription model, you have voluntary churn (where the user chooses to leave) and involuntary churn (where the user doesn’t mean to leave, but churns because their payment fails).
Just about every single payment system, out of the box, offers their Retry Cycles and the Payment Failure emails, especially if it’s a soft decline that the user is churning on. But I’m here to tell you: That. Is. Not. Enough.
The importance of in-product messaging for this stuff cannot be overstated. At Lovable, we’d just been doing the emails, but finally put the messaging in the product to say, ‘Hey, your payment has failed.’
Do you want to guess how much this improved our recovery rate (relative)?
50%. Fifty f^%*-ing percent. No joke.
It’s just absolutely wild. But it makes sense: These people clearly didn’t realize this was happening. Email is great for some things, but you can’t always depend on it. You need to have something in-product telling people. We went from ~20% of people making it out of ‘Payment Failure Retry’ updating their payment method to >30%. This is huge.
And to be clear: The incrementality of this is questionable, because if people really want to stay… even if they churn for a bit, they’ll be back. They’ll just resubscribe when they realize it. But there’s no reason for this revenue lag for you!
Fixing this also improves the overall user experience, because if they wanted to stay and didn’t mean to churn, then notifying them of the potential failure helps to prevent confusion.
If there’s an easy win, this is it. Go make sure that any payment failures are communicated in the dashboard.
#3 - 5 free daily credits for everyone
This one is specific to credit-based monetization models, but giving away a small number of daily credits to keep people engaged is incredibly smart.
Lovable was already doing this before I joined, and it’s genius. Free users get 30 credits per month, but can only use 5 per day. And even if a paid user runs out of credits, they still get 5 free credits a day on us.
It’s a very smart mechanic because it keeps people active, keeps them coming back, and reduces the chance that hitting a limit turns into abandonment.
I’d particularly encourage AI companies to explore this, since the go-to approach right now tends to be gating everything. ‘Want some AI credits? Maybe a quick trial, but everything else you’ll pay for.’ This may get people in the door, but it doesn’t keep them around.
Instead, this is an ongoing hook. ‘Your 5 credits refresh tomorrow, you should come back and use them!’ works super well. People are waiting with excitement for those credits to refresh… and often they realize that it’s not worth waiting, so they decide to just go ahead and buy.
#4 - User-friendly policy for credit rollovers
This one is so powerful. Mostly because of how most companies in this space operate. You often buy AI credits and then they just expire at the end of your subscription term if you don’t use them.
Obviously, this makes for very easy accounting. But it’s sooo counterintuitive for a user.
For AI products especially, usage is very bursty. It goes up and down. Some up months or weeks, I’m very active. Other days or weeks, I might be completely inactive. If the company expires credits that people have purchased too fast, people get pissed.
If they didn’t use the credits, the users (understandably) become resentful that you took away something that they paid for. This is not platform or per-user pricing! ‘I bought the credits. Why did they leave? Like, where did they go? I didn’t use them. What do you mean I’ve lost access to them.’
So, the way we’ve structured this is that users get to rollover their credits, as long as they’re on a paid plan. If they fully churn, those extra credits are frozen until they resubscribe.
Not surprisingly, the likability of the product has increased dramatically, because people appreciate that we’re accommodating the bursty usage, as opposed to snapping them into a very harsh use-it-or-lose-it subscription model.
A bit more surprising though is that this has also improved churn. People see that they have all of these credits accumulated in rollovers, so they end up coming back. There’s a natural desire to use up their accumulated credits before churning, so they don’t miss out on what they already have. And then, once they start building again, they realize how great it is and catch the bug to keep going. So it’s a very natural trigger: using something that they’ve already paid for, and just holding it for them. They’re happy. They’re engaged (or re-engaged). Everyone wins!
#5 - Top-ups.
I already did a full post about this, but I want to call it out as a relevant tactic.
(Check out that post: We stopped forcing the subscription model on our users. Here is what happened. - How and why Lovable introduced ‘top-up’-style monetization)
This can really move the needle in terms of retention. Maybe it sounds crazy that just adding an ad hoc payment option to your subscription-based model could really change things, but it did for us. I’m talking a 7% improvement to our overall retention! That’s no small feat when you’re talking about a base of a million paid users at Lovable.
But the point here isn’t necessarily that you need to do top-ups specifically… the point is that people were having frustrations with our monetization structure. By introducing the framework that they naturally wanted to use to transact with us, we created a better overall experience with us, which made them want to stay.
2 Laws
I feel like I’ve said these over and over and over, but I keep seeing people not embrace them.
1 - Retention tests take time.
In order for us to observe the impact of all the tactics I mentioned above, we had to observe the tests for multiple months. This is not something you can see in a week.
Let me be as clear as I can about this: When you’re doing retention drivers, you often see an immediate negative impact on revenue. But then, in a month or two, the revenue’s coming in hot and everyone’s all ‘Holy shit, this is amazing, more of this please.’ It ends up being absolutely obvious net positive… if you wait long enough. But so many times, people react negatively too quickly.
2 - Retention tests often feel counterintuitive.
Retention depends on user value. And user value often has to be paid for at the cost of (short-term) company value.
Teams often struggle with this… You’ve got to give more, often giving stuff away for free! In some ways, you’re optimizing exactly against your monetization model, not for it… and yet, that makes your monetization overall much more healthy. This is one thing I see that separates decent Growth teams from really incredible Growth teams - don’t just optimize your pricing. Optimize the entire customer experience, because that’s what ultimately translates to retention and long-term monetization impact
All of these changes came with a lot of anxiety - it’s not easy to hold out when something doesn’t look good initially, or to optimize against your own model. But it’s paying off. Literally.
So be patient. I know our industry doesn’t value this or reward it very often. But if you want your users to stick around for the long term, consider how your actions will impact the customer relationship over that timeframe. And try measuring the impact of what you’re doing beyond a one-week test. Incredible retention is worth the wait.
Edited by Jonathan Yagel.