How to Kill Churn: Everything I Learned in 10 Years.

After scaling VEED to $50M in annual revenue in last 8 years, I’ve seen churn from every angle: Self-serve mobile, Enterprise and API.

Today, I'm sharing everything I’ve learned (the hard way) about churn and how you can solve it in your business👇

1./ Churn depends on your category.

It's about the market you picked, not the product you built.

Build a tool a marketer uses every day → low churn.

Build a tool someone uses to make a birthday card → high churn.

– Enterprise products sit inside someone's daily workflow, so they retain.
– SMB churns higher because small businesses go in and out of business.
– Consumer churns the highest because the use cases are often one-time.

If you picked the market well, the rest is optimization. If you pick it wrong, no amount of exit flow magic saves you.

The benchmarks I use:

And the math that should terrify you:

10% monthly churn = you replace 70% of your user base every single year.

At $1M ARR, no big deal.

At $100M ARR, you need $10M in new ARR every month just to stay flat.

But if you have hundreds of millions of users (Spotify, Tinder, Bumble), high churn is fine since the reactivation and top of funnel is massive.

2 more terms worth knowing:

The uncomfortable truth: you can optimise churn. You can't dramatically move it.

- At 20%, you're not getting to 3%.
- At 8%, you'll be lucky to get to 3%.

Over 3-5 years of focused work, maybe you can cut it in half. That's it.

2./ Your blended churn number is a lie.

Not all churn is bad. Some is fine. Some is killing your business.

You can have a blended churn rate of 10%, but inside that number, some cohorts churn at 5% and some at 20%.

The 20% is what's dragging your average down.

The whole game is this:

> Find the sticky users
> Build for them
> Acquire more of them

- Step one is segmentation.
- Add an onboarding survey.
- Ask role, company size, use case.
- Then map that data against churn.

You'll start to see who your stickiest users actually are and who's churning out fastest. Step two is talking to these sticky users. Properly.

I personally onboarded customers until VEED was at $6M ARR. I spoke to 10-15 people a day. I barely looked at the data. I just listened.

The trick: give value first.

I once sent an email asking users for feedback. Got zero replies.

Changed the subject to "I'll personally onboard you onto the product." My calendar filled up immediately.

Most founders skip this work because it feels less impressive than redesigning the homepage. Do it anyway.

In the early days of VEED, I'd get videos from a kid in the US making cops-and-robbers school projects. Twenty minutes later, I'd jump on a call with a marketer making one video a day for social. Low churn. High LTV.

My job then became to acquire fewer kids and more marketers.

A reality check: shifting your user base takes 12-24 months. It's slow. It's brutal. And it's much easier early than at scale. So fix it early.

3./ Activation is your secret weapon.

Once you know your sticky ICP, the product job becomes simple: build for them, ruthlessly.

In every business, 2-3 things drive 60-70% of usage. Find them. Go deeper there than anyone else.

Then there's the most underrated churn lever of all: activation.

If 10 users sign up and only 1 hits the meaningful action, you have a "terrible" activation rate.

Find how long it takes a new user to hit the aha moment. Friction kills it.

Credit card details upfront.
Complex UI.
Long marketing surveys.

All of these destroy time-to-value.

Which brings me to the free plan.

Your free plan = Your secret activation tool.

A good example is Cursor. You start building, you lock in, and soon you realise you are out of credits. Then you end up paying because of the sunk cost fallacy. Note: this is completely opposite to the strategy Duolingo uses.

4./ The 3 hidden levers you aren't using.

Some churn levers are tactical.

Use them to take 10-30% off your churn.

1. Annual plans. Contractually locks users in for 12 months. Churn drops mechanically. You also get a year's revenue upfront. Watch the discount too. 80% off annual means a company has a really bad churn problem. 20-40% is the healthy range.

2. Exit flows. Use a tool like Churn Key. Offer pauses. Offer 3 months for $1. Ask what feature was missing, then go build it. I signed up to Paramount, tried to cancel, got 3 months for $1, binged Land Man in 2 days. I'm probably still paying them today.

3. Delinquent churn recovery. Up to 2% of churn is just expired cards. Set up Stripe smart retries and dunning emails to recover it automatically.

Then there's the holy grail: negative churn.

Negative churn means the customers you keep grow faster than the ones you lose. Your existing accounts add seats, upgrade plans, or buy more products.

If Figma stopped acquiring new customers tomorrow, its revenue would still grow. That's negative churn.

How do you get there?

Also, here's how I rank businesses on the basis of churn.

TL;DR

Here is a video walkthrough for those who are REALLY into churn - youtu.be/vdCi2GAQA_Y?si…