Prices
What a custom Shopify app costs to build and run
Most pages on this question answer it with a range so wide it is useless. Here are the actual numbers I charge, what moves them, and the ongoing costs that quotes routinely leave out.
Checked 30 September 2026
What I charge
| Build | To build | Monthly after |
|---|---|---|
| Custom subscription app | £1,000 | £50 |
| Inventory and fulfilment system | £1,500 | £79 |
| AI customer service desk | £1,000 | £69 |
| 30 SEO landing pages | £1,000 | none, nothing to run |
| Something else | Quoted before work starts | Depends what has to run |
Fixed before anything is written, not an estimate that moves. The monthly exists where I run something on my infrastructure and does not exist where I do not, which is why the SEO pages have none.
Why quotes for the same thing vary so much
Who is doing the work. An agency carries salespeople, project managers, account managers, an office and a marketing budget, and all of that is in the price before a line is written. That is not a criticism, it is what buys you capacity and cover when someone is ill. You are choosing whether you need it.
Day rate versus fixed price. A day rate moves the risk of the estimate being wrong onto you. A fixed price moves it onto the person quoting, which is why fixed prices come with a scope document and why changing your mind mid build costs money. I quote fixed, and I would rather spend an extra day on the scope than discover the gap in week three.
Scope creep dressed as discovery. A long paid discovery phase before anyone commits to a number is sometimes genuinely necessary and is sometimes a way to bill you while deciding whether to bid.
The costs quotes leave out
Hosting. An app is a server and a database and they cost money every month for as long as the app exists. If a quote has no ongoing figure, ask where the app is going to run and who is paying for it.
Shopify's API changes. Shopify versions its API quarterly and deprecates older versions. An app nobody maintains stops working, not immediately but reliably. Budget for someone keeping it current or accept that you are buying something with a shelf life.
Monitoring. Somebody has to notice when it breaks. Preferably before your customers do.
Changes. Your process will change within the year. If every change is a new quote, the cheap build is not cheap.
Working out whether it pays back
The arithmetic is simple enough to do on the back of an envelope. Take what you currently pay for the app you would be replacing, including the percentage and per order fees. Subtract the monthly on the build. That is your saving per month. Divide the build fee by it and you have the number of months to break even.
On a subscription app at £40,000 a month of subscription revenue, a rented app on a percentage runs to several hundred a month. Against £50, the £1,000 build is paid back inside a quarter. At £5,000 a month of subscription revenue the same sum says do not do it, and that is the right answer.
What the arithmetic does not capture is the process you are currently doing by hand because the app cannot. That is usually the real reason people build, and it is worth more than the saving.
What a fixed price means here
The number is agreed in writing before anything is written, along with what it covers and a date. If the build takes longer than I expected, that is mine to absorb. If you add something substantial partway through, it is quoted separately and you decide before I start it.
Paid up front, and refunded in full if I cannot deliver what was agreed.
Get the arithmetic on your store
Send me the apps you pay for and roughly what you process a month. You get back what a build would cost against what you spend now, including when the honest answer is that it does not pay back.