The Subscription Price Increase Math Most Founders Won’t Run
TL;DR
- Netflix and Spotify both raised prices in early 2026. Subscribers stayed. The "price increase kills retention" story is folklore.
- A 15% increase can lose you 8% of the base and still grow monthly revenue. The math is less scary than the email.
- The churn number that spikes after a hike is the wrong number. Cohort revenue is the right one.
- The rollout costs more subscribers than the increase. Five rules for sending the email without wrecking the base.
Two truths, one spreadsheet 📉
Your CFO says raise the price. Your retention lead says we'll lose 8% of the base. They both believe their math. Only one of them ran it.
Here's the reality: most subscription founders would rather quietly bleed margin than send the email that says "your price is going up." The email feels like a betrayal. You spent years earning that subscriber, and now you're asking for more.
The data on what actually happens is a lot less scary than the story in your head.
What Netflix and Spotify just did 📈
In March 2026, Netflix raised prices across every US plan - standard from $17.99 to $19.99, premium to $26.99. It was the first hike on their ad-supported tier, ever. They had 325 million subscribers at the time.
In January 2026, Spotify lifted US Premium from $11.99 to $12.99. In the UK, it was their third increase in two years.
Neither platform collapsed. The subscriber exodus every Reddit thread predicted didn't show up. Netflix has lock-in you don't, so don't copy them blindly. But the takeaway generalizes: price sensitivity is far more context-dependent than founders assume. The same person who rage-posts about a $2 increase keeps paying for three streaming services they barely watch.
Run the math
Let's make it concrete with a worked example. Swap in your own numbers.
Say you have 10,000 subscribers paying $20 a month. That's $200,000 in monthly revenue. Assume 60% gross margin.
You raise the price 15%, to $23. Your retention lead's nightmare scenario: 8% of the base churns specifically because of the increase. That's 800 lost subscribers.
The surviving base: 9,200 subscribers at $23 = $211,600 a month. You lost 800 people and your revenue went up $11,600.
On margin it's the same story. Gross profit before: $120,000. After: $126,960. Up roughly $7,000 a month, every month, compounding.
Most people stop the math here.
The 800 who left were your most price-sensitive subscribers, the ones with the weakest relationship to your product. They were going to churn anyway, sooner, for a cheaper competitor or a free alternative. You accelerated churn you were already going to eat, and you raised the value of everyone who stayed.
The number that's lying to you 🎯
Here's where founders panic: they watch the headline churn rate tick up the month after the increase and conclude it failed.
That number is a lagging, blended mess. It mixes price-churn with seasonal churn, with the people who were going to leave anyway, with the noise of a single billing cycle. A 1.5 point bump in churn looks alarming in a dashboard. In cohort revenue it can be invisible.
Track the cohort that got the increase. What's their 30-, 60-, 90-day revenue per surviving subscriber compared to the pre-increase cohort? That's the number that tells you whether it worked. Headline churn tells you whether your retention lead is upset.
The fix: how to roll it out 💡
Here's how to send the email without wrecking the math:
- Grandfather the loyal base. The people who've been with you 12+ months are your highest-LTV subscribers. Hold their price while new signups pay the new rate, and let annual renewals migrate over 6 to 12 months. Reward the loyalty, don't tax it.
- Sell the value you've added. "Due to rising costs" is an apology. "Here's what we've shipped since you joined - three new features, faster support, the thing you asked for" is a reason. Subscribers accept an increase when they can see what they're getting. The moment it reads as margin defense, you lose them.
- Give 30 days notice, minimum. The email costs you one uncomfortable day. A surprise increase on an annual plan costs you a quarter of chargebacks and bad reviews. 60 days is better. The notice is the retention move.
- Segment by risk. Don't blanket-increase the entire base at once. Start with a cohort whose usage proves they can't easily leave, or with short-tenure subscribers who have the least entrenched price expectation. Test, measure the cohort delta, then expand.
- Offer the downgrade exit. Put a cheaper tier or a pause option in the same email. Some subscribers will take it. A downgraded subscriber is worth less than a full-price one, but vastly more than a churned one. The exit ramp is retention.
The honest caveat
Run the math, then decide. The math can tell you not to raise - if your value perception is weak, you just shipped a buggy release, or your last three months of product work have been invisible to subscribers. A price increase on a product that's quietly decaying will accelerate the decay. The increase amplifies whatever trajectory you're already on.
And the 8% churn assumption in the math above is a guess, not a law. Your real number could be 4%. It could be 12%. The only way to know is to test on a small cohort and measure. What the math does is simpler: it shows the bar for "raising prices is worth it" is much lower than your gut thinks. It won't guarantee the outcome.
After 11 years running retention at Scentbird, the pattern I saw was consistent: the teams that ran the math raised prices on a cadence and reinvested. The ones that didn't quietly absorbed the margin hit year after year until it became someone else's problem. (This is a big part of why we built Finsi - to make the cohort math obvious enough that the right call stops feeling like a bet.)
So here's the question worth sitting with: if you raised your price 10% tomorrow and lost 5% of your base, would your business actually be worse off, or would you just feel worse?