Dose 220: Is Your Subscription Churn Good or Bad? Here's My Benchmark

This week’s dose is also a full podcast interview with Varun Kunrda about making the most of post-purchase upsells. Tune in on your favorite platform:

Quick Note on ESPs

Here's how I'm thinking about ecommerce tech stacks right now, especially email. The brands pulling ahead are getting better ROI from more affordable systems, testing faster with AI, and working with software partners that actually show up.

Our partner Omnisend checks all three: up to 35% less than Klaviyo, AI built in at no extra cost (you can manage the whole platform from Claude or ChatGPT), and real humans answering support in 4 minutes. They'll even migrate you free in 5 days.

Get an additional 30% off when you signup with code SUBDOC30 at signup.

Matt here with your weekly Subscription Prescription 💊

This week I want to answer the question I get on almost every call, client or not: is my churn good or bad?

It sounds simple. It rarely is. And I have a pet peeve about how it usually gets answered.

The big subscription platforms publish benchmark reports every year. You've seen them. Bands by revenue, bands by vertical, nice charts. The extremes are easy to dismiss. The middle is where the confusion lives.

If a brand doing a billion dollars has 4% churn and you do $10,000 a month, you know that number doesn't apply to you. But when you do $10 million and the brand next to you does $15 million in a different vertical? The report can't tell you anything useful about that gap.

Context is everything with churn. So today I'm giving you how I actually evaluate it when I pull up a program.

1. Single digits is the baseline, and growth rate changes the grade

Here's my starting point. Your average monthly churn should be in the single digits.

The math is simple. Start the month with 1,000 subscribers, add 100, lose 100, and you end at 1,000 subscribers with a 10% churn rate. That's your aggregate number, and it's the first one I look at.

If you're at 15% a month, that's not good. Anything above it is really bad. You can run a program at 15%, but it's hard to scale, because you replace a big chunk of your base every month just to stay flat.

Now the caveat, and it matters. How fast you're growing changes how I read the number.

If you're a hyperscaler, meaning ad spend and new subscriber counts climb every month, you reach a broader and broader audience. That doesn't mean your offer is perfect for your ideal customer. It means it works for them and for everyone adjacent, and some of those people were never going to stick.

So a fast grower at 12% does not worry me. That can scale for a while, as long as you know you'll have to dial retention in eventually.

But if you're not a hyperscaler and you're at 15%? Something in the program is off. That's a diagnosis, not a benchmark.

Vertical matters too. Skincare brands with a product that actually works can run under 5%. Think about how a Dermalogica or True Botanicals customer uses the product every day. Supplements usually sit higher. Meal plans run high unless the offer is extremely dialed in.

Takeaway: Pull your aggregate monthly churn, then grade it against your growth rate and your vertical. A 12% number means two different things at two different brands.

2. The monthly average hides where churn actually happens

The aggregate number is useful, but blunt.

A subscriber nine months in churns at a fraction of the rate of someone in month one. So I look at churn as a time series: month one, month two, month three, and so on. That's where the real story is.

And inside month one, I go one level deeper. What does churn look like in the first seven to ten days?

That window is special. Nobody has prompted those people about a renewal yet. There's no billing reminder, no upcoming charge, maybe an onboarding email or two. If they cancel within a week, that decision came from them alone.

If that early number is more than a couple of points, one of two things is usually going on.

Either the offer is too good to be true, and people grab the discount with no intention of staying. Or the onboarding isn't doing its job and people are confused before the product has even arrived.

Churn at the first renewal tells a different story. If people bail right when the second charge comes, I look at the offer from a different angle. If the first box is realistically a two or three month supply and you're renewing monthly, of course they're canceling. They don't need more yet.

Grüns handles this well. The first order is sized to the cadence, so the subscriber runs out about when the next box lands.

Takeaway: Build a cohort view of churn by month, and break month one into the first ten days and the renewal point. Each window points to a different fix, and the aggregate won't tell you which one you need.

3. Set the next goal, not the dream goal, then pick the right lever

This is where I get on my soapbox a little.

If you sit at 15% monthly churn, you should not be targeting 5%. Target 12%. Then 9%. If you’re at 10%, let's get to 7%. At 5%, I'm aiming for four and a half, maybe four and a quarter.

Big jumps don't happen. Incremental ones do, and they compound. Realistic goals also keep the team from chasing a number your category was never going to hit.

Which lever you pull depends, again, on growth rate.

If growth is strong and churn is around 10 to 12%, leave the offer alone. It's working. Spend your time on the post-purchase side: onboarding, billing reminders, cancel flows, dunning, win-backs. Each of those can shave a point. Twelve becomes eleven, eleven becomes ten.

If growth is slower, your offer structure is on the table. Bulk subscriptions, starter kits, anything that brings the right type of person into the program instead of the most people. That usually means taking a little less off the top and being less aggressive about who you pull in.

Here's the trade I'm happy to make almost every time. If giving up 10% of the subscribers I acquire gets the ones I keep to last 10% longer, that's a win. Fewer, better subscribers compound into more revenue and profit than a bigger, leakier base.

Takeaway: Set a churn goal one step from where you are today. Then decide whether the fix lives in retention or in the offer based on your growth rate, not on what the benchmark report says.

Bottom line: No single number tells you whether your churn is good. Single digits is the baseline, but your growth rate, your vertical, and where in the lifecycle people leave all change the answer.

Pull the aggregate, then the cohorts, then the first ten days. Set a goal you can actually hit this quarter. And if you want a second set of eyes on your cohort retention or your offer, reach out. It usually doesn't take long to spot the leak.

Until next Tuesday, that's your Subscription Prescription. 💊

- Matt Holman 🩺