You can raise your price tomorrow and not one existing subscriber pays a cent more. Every price you ever shipped is still live in your billing system, charging whoever signed up under it. Grandfathering forever is a decision, and teams usually land there by accident.
Showing โฌ43.21 on your pricing page and then charging in USD is the worst version of localized pricing. The customer's bank runs its own conversion at its own rate and adds a 1-3% foreign transaction fee, so their statement shows a third number you never quoted.
"You will not be charged today" is technically true and it still fills your support inbox. A pre-auth hold shows up in the customer's banking app as a pending line with a real dollar amount on it. Say what they will see, and when it clears, on the page.
Most founders picked their free plan because a competitor had one. Almost nobody runs the two numbers that decide it: what a free user costs to serve per month, and how many paying users they bring in. If it is $3 and zero, that is ~$36 a year each for a demo that never closes.
Run one query on your last 1,000 trials: how many are the same person back with a new email address. Most billing dashboards have no field for it. Plenty of founders know their CAC to the cent and their repeat-trial rate not at all.
The customer who quietly stops using your product is a churn event that has not been recorded yet. Usage going to zero is the earliest signal you get, and it arrives weeks before the cancellation does. Almost nobody watches it, because it does not show up in revenue until later.
Discounting to save a cancellation usually just moves the cancellation. The customer who was not getting value at full price is not getting value at half price either, they are getting it more slowly. Fix the reason they were leaving or let them go cleanly.
The cancel flow is the last place most teams put effort and the first place they could win. Not a retention wall, just one honest question about why, and one relevant alternative. A downgrade you offered beats a cancellation you argued with.
Annual plans do not reduce churn, they delay it and hide it. The customer who would have cancelled in month three cancels at renewal instead, and you had eleven months of no signal. Watch usage on annual accounts or you find out on the day the money stops.
Your refund policy is a retention tool and most founders treat it like a legal document. A fast, no-argument refund turns an angry churn into a neutral one, and neutral customers come back. The fight you win over $47 costs you the second subscription you never hear about.
Your pricing page gets blamed for conversion problems your checkout caused. People decided to pay, hit a card form with nine fields and a surprise account-creation wall, and left. Count the drop-off between 'clicked buy' and 'payment done' before you touch pricing again.
Not all failed payments are equal. A hard decline (closed account, stolen card) is dead, stop retrying. A soft decline (insufficient funds, a temporary hold) often clears if you retry at the right time. Treat them the same and you either lose real revenue or hammer a dead card.
Most failed-payment emails fire at the worst moment: instantly, before the bank glitch that caused it has even cleared. Wait a few hours, retry, then space the next tries over days. A smart retry schedule recovers more than the angriest 'update your card' email ever will.
The cheapest retention win nobody sets up: fix the card before it fails. Card networks push updates when a card is reissued or the expiry changes. Catch that update and the charge just goes through. Skip it and a happy customer churns over a number that changed.
Most founders can quote their cancel rate to the decimal and have no idea what share of it is just failed card charges. A card expires, the retry gives up after one try, the customer never meant to leave. That's not churn, that's a billing bug wearing a churn costume.
55M emails and tens of thousands of Stripe payment records, partial cards included: that's the Suno breach. Most SaaS teams file it under someone else's problem, then wonder why their trial funnel keeps filling with stolen cards. Gate the card at signup.
Most founders price off the number they want to net. Stripe's standard rate is 2.9% plus 30 cents per charge, so to actually keep $100 you have to bill $103.30. Almost nobody runs that math backwards, and it quietly eats a few points off every plan.
Trial farmers cycle through disposable emails but reuse a small set of real cards. Stripe gives every card a fingerprint that stays the same across accounts, so a card flagged on one trial signup can be caught on the next one, even behind a brand new email address.
Trial signups are a trailer metric: mood, curiosity, burner emails. Revenue is opening weekend. The Odyssey just proved it: 700,000 trailer dislikes, then the best audience score of Nolan's career once it shipped. Pre-release numbers are noise. Shipped numbers are money.
Big week for stolen credentials. The legal version gets no headlines: a temp email and a prepaid card mint a brand-new "customer" in most free trials on the internet. Founders eat it on purpose. Measuring trial abuse would shrink the signup chart investors see.