The How to series · Post 22
The refund promise: why we answer for the work with money
For UK business owners deciding whether to trust another AI supplier: what the trial-and-hope cycle already costs you, and the promise we put our own £499 behind.
An owner reads another website promising hours back. They remember the planning tool from two years ago, bought on a demo and abandoned by spring, still billing monthly until somebody noticed. Nobody rang to ask what went wrong. The tab gets closed.
That's the loop this post is about: buying fixes that don't stick. Not because every tool is bad. Because the risk sits entirely with the buyer. You pay first and you do the set-up, and if it doesn't hold, the loss is yours while the supplier's site keeps selling the same promise to the next person.
We have no clients. Zero. Pre-launch, by choice. So this post carries no testimonials and no case studies, and we won't invent any. What we have instead is a promise with money behind it: the £499 diagnostic finds 5 hours a week worth automating, or you get the full refund. No small print, no negotiation. If a build follows, the £499 is credited in full against it. Either way, the risk sits with us, not with you.
Why answer for the work with money? Because the alternative is asking you to believe a website, and you've met those websites. Our founder, Pete, has spent 25+ years in tech building for other people's businesses, and that history taught him what a promise is worth on its own: very little.
The promise is not free for us to make. Every diagnostic is real hours before any build exists. If the hours don't turn up, the fee goes back and our time is already spent. We priced that into how we work, not into small print.
The maths (a sketch, not a client figure)
The trial-and-hope cycle, priced with assumptions you can edit:
- Two tools a year tried and quietly dropped (count your own; most teams recognise this)
- × £50 a month each, kept six months past the point everyone knew it wasn't working = £600 a year
- × ten hours per tool getting the team onto it and off it = 20 hours at £30 an hour = £600 a year
About £1,200 a year on fixes that didn't hold, before counting the goodwill lost inside the team on another false start. Swap in your own numbers. The shape rarely changes, and it explains why "trust us" is a hard sell in this category.
Score it: the three checks
Score the loop "testing a fix before money moves":
- Hours. The evaluating and unwinding each tool costs, spread thin but real: 2.
- Repeat. Could a new starter follow "how we test a fix in one week" from one page? Yes, if someone writes it: 4.
- Risk-if-wrong. A dropped test wastes money but burns no client, and you see it quickly: 4.
Verdict: two at 4 or more, start narrow. Write the one-page test and run it on the next tool before you commit. Then hold every supplier to it. We hold ourselves to the same shape: a test with a fee attached, refunded if the hours don't turn up.
Try one thing this week (free)
List your last three software or AI purchases. Next to each: what it promised back then, and the hours it actually saved. Ten minutes from memory. Judge it on Friday: if you can't name the hours saved on at least one of the three, you're buying on hope, like most of the market.
Where a person stays
The promise itself is approval-first. The diagnostic's 5 hours are measured with you on a call, and the ranked report is written by a person whose name is on it. Nothing is claimed that you can't check against your own week. The same rule runs through anything we build: software prepares and a person decides. Nothing sends unchecked, and anything touching money, compliance or legal stays approval-only.
FAQ
You shouldn't, not on words. That's why the answer is a refund, not a claim: 5 hours found or the £499 comes back, and the £499 is credited in full against any build that follows. The risk sits with us.
A one-hour call, a ranked written report, and a 30-minute follow-up where we recommend what to build first. If fewer than 5 hours a week turn up, full refund.
We don't have any, and inventing them is the oldest trick in this category. The promise is the proof we can stand behind today. Testimonials arrive when they're real.
Real hours before any build exists. If no hours turn up, the fee goes back and the time is spent. We'd rather carry that than ask you to carry the risk.
No, and we say so plainly. The promise covers the diagnostic. A build is priced separately, after a diagnostic, and the £499 comes off that price in full.