Custom software earns its place in a small business when one process is costing you real hours every week and nothing you can buy fits the way you actually work. Below that line, buy off the shelf. Above it, building usually wins, and it wins by starting small rather than by starting big.
Most guides on custom software development for small business buyers in the UK skip the part where the answer is "don't build". This one does not. Here is how to work out which side of the line you sit on, what UK firms actually charge, and what the payback looks like when it lands.
What counts as custom software for a small business?
Software built around one company's process rather than sold to a market. In practice that means internal platforms, job sheet and workflow systems, quoting and invoicing tools, customer portals, and the integrations holding your existing tools together. It rarely means an app in the app store sense. Most of what we build for smaller UK firms is a web system used daily by a few office staff and a field team on their phones.
The distinction that decides the money is not custom against off the shelf. It is whether the process is an asset or a commodity. Payroll is a commodity, so buy it. How you price a job, schedule your crews and prove the work was done may be the reason customers pick you over the firm down the road, and there is often no product on the market that matches it.
Is custom software worth it for a small business?
It is worth it when you can name the hours it gives back, in advance, without squinting. The UK government's SME Digital Adoption Taskforce final report (July 2025) puts the productivity gain at 7 to 18 per cent per technology adopted, a figure it attributes to the Enterprise Research Centre, and cites Be the Business in estimating that a 1 per cent productivity uplift across UK SMEs over five years would add £94 billion a year to GDP.
National figures do not pay your invoice. The test that does is narrower, and you can run it on the back of an envelope. Take one process. Count the hours it eats a week, multiply by a loaded hourly cost, multiply by 48. If that annual number is well below the cost of a build, buy something or leave it alone. If it is a multiple of the build cost, you have found your first project.
The same report is clear about why more small firms have not run that sum. The UK sits 25th in the world for future digital readiness on the IMD 2024 index, UK SMEs invest less in technology than their G7 peers, and 43 per cent of SMEs surveyed in 2024 had no plans to use AI at all, according to British Chambers of Commerce data quoted in the report. The barrier is time, cash and knowing where to start, in that order.
How much does custom software cost for a small business in the UK?
Published UK bands run from roughly £20,000 to £200,000 depending on who you ask and what you count, with most small-business systems at the lower end of that. Two agencies who publish their numbers give a reasonable picture of the market:
Treat those as market anchors rather than quotes. Our own numbers come out of scoping a specific problem, not off a price list, and any agency quoting you a figure before understanding your process is guessing. We wrote about why identical briefs attract quotes that differ by a factor of ten in what it actually costs to build an app.
The more useful number is the first one. A first phase that solves one process, ships in weeks and earns its keep is a far better bet than a platform that does everything and lands in a year. Sequencing beats scale here, and it is the cheapest risk control available to you.
What does the payback actually look like?
On the work where it lands hardest, the change is in the order of magnitude rather than the percentage. Two of ours, both real and both published on our own site: an invoicing cycle that ran to six weeks now completes in six hours, and an expense claims process that handled 50 claims a week now handles 1,500.
Neither of those came from clever technology. They came from removing the retyping, the chasing and the spreadsheet that three people maintained different versions of. The pattern repeats: the value sits in the handoffs between people, not in any single task.
When should you buy off-the-shelf instead?
Buy when your process is ordinary and a product already fits it. Accounting, payroll, email, CRM at the simpler end and document storage are all commodities with mature products, and building your own version of one is an expensive way to end up behind.
Buy, too, when you cannot yet describe the process. If four people in the business would explain the workflow four different ways, software will freeze that confusion in code. Fix the process first, on paper, then decide whether it needs building.
The middle ground worth considering is a no-code tool for the pilot. It gets you something usable quickly, and it hits a wall on ownership, integration and scale, which we set out in no-code got you a demo. Use it to prove the workflow, plan to replace it if the business depends on it.
How do you start without betting the business?
Scope one process, fund one phase, and set a number you expect it to move before anyone writes code. Our own discovery process exists for this: work out what is worth building, then build that. Roughly half the value of the exercise is the features it talks you out of, which is the argument in how many features an MVP should have.
Phase one should be small enough that you would survive it failing, and specific enough that you would notice it working.