If you have fewer than roughly 2,000 paying customers, do it yourself in a spreadsheet. If you have more and want a number first, a one-time report is the lowest commitment. If you already know what to fix and want it run, you need retention software or a service. Here is how I would choose between these three, with the costs and limits of each.
What are the three options?
- DIY spreadsheet: export invoices and subscriptions from Stripe and count the leaks yourself.
- One-time report: someone reads your Stripe data with a read-only key and gives you the leaks in dollars. This is the Revenue Leak Report, $490.
- Retention software or service: something that sends messages and offers to your customers and measures the result.
I sell the second and the third, so read my comparison with that in mind. I will tell you where the first is the right choice.
When is the spreadsheet the right choice?
Below roughly 2,000 paying customers. There are too few events in each bucket for a paid analysis to say much, and the cost of the work is better spent elsewhere. I say this to people who ask me directly, and the landing page says it too: if you are smaller and unsure, ask and I will tell you before you pay.
A spreadsheet needs a year of invoices, cancelled subscriptions and trial data, and the counting rules written down. The rules are the hard part, not the arithmetic. My own are in what involuntary churn is and how to calculate a failed-payment recovery rate. You are welcome to copy them.
When is a report better?
When you have the volume, and the counting across thousands of invoices is the slow and error-prone part. The report does it from a restricted read-only key in about an hour of my work and 48 hours of calendar time. It gives you four leaks in dollars per month, the recovery rate, the revenue at risk now and the top three fixes, with a method page that lets you check it.
It also suits you if you are not sure you have a retention problem. $490 is a cheap way to find out which of the four buckets is large and which can be ignored. See is the $490 report worth it.
When do I need software or a service?
When the numbers show a leak large enough to justify action, and you do not want to build and run the messages yourself. That means failed-payment sequences, cancel flows, trial nudges, plus the measurement. What matters is how it is priced and how it is proved.
Questions I would ask any provider, including me:
- How do you prove the result? Is there a control group that gets nothing, or is it a before-and-after chart?
- What do I pay if nothing improves?
- Do messages go out under my brand and domain or yours?
- Can I see and undo every change?
- What access do you need to my Stripe account, and for how long?
How does my pilot answer those?
A 90-day pilot costs $1,500 setup plus 20% of measured net uplift. The treated group is measured against a 20% holdout that gets nothing from me, and uplift is revenue times margin minus the cost of the offers. With no uplift, there is no percentage. I send from your own brand and domain, every change is recorded as a request that can be undone, and any offer that costs more than it can win back does not go out. Email is live; WhatsApp, SMS, Viber and Telegram are available on request and are not part of the standard pilot.
The details are in how a pay-on-uplift retention pilot works.
What does each option not give me?
- Spreadsheet: no fresh pair of eyes, and nobody to catch a counting error. It also takes your time.
- Report: no proof that a fix works. It is a read of history. It also cannot see saves made outside Stripe.
- Software or service: a longer commitment and more access. It is only as good as the measurement behind it.
Can I combine them?
Yes, and I would. Start with the spreadsheet or the report to learn the size of each leak. Fix the cheapest one yourself if it is simple, such as turning on the failed-payment emails you may already have. Bring in a service only for the part where you want the measurement and the sending done properly.
I would not buy software before knowing the numbers. Without them, you cannot tell whether it paid.
What is the cost of waiting?
The leak keeps running. If the realised leak is a few thousand dollars a month, each month of delay costs about that much, and a cheap measurement now is worth more than a perfect one later. If it is small, waiting costs little, and you should spend your time elsewhere. That is the logic behind measuring first.
How do I decide in five minutes?
- Do you have roughly 2,000 or more paying customers on Stripe? If not, spreadsheet.
- Do you know your leak in dollars per month? If not, report or spreadsheet.
- Is the largest leak one you can fix with a short sequence of messages? If so, try it yourself with a holdout.
- Do you want it run and measured for you? Then a pilot, after the numbers.
The report is $490, delivered within 48 hours of my receiving a restricted read-only Stripe key. You can look at a sample with synthetic data first.
Want your own numbers?
The Revenue Leak Report puts failed payments, unconverted trials, cancellations and downgrades in dollars per month, read from a restricted read-only Stripe key. $490, delivered within 48 hours of receiving the key. Nothing in your account is changed.
Get the Revenue Leak Report - $490 See a sample (synthetic data)