If a product already does what your business needs, buying it is nearly always cheaper, quicker and safer than building your own. Bespoke software only makes sense when nothing on the market fits, or when the software is part of what you sell. That may sound like an odd thing for a company that builds bespoke software to say, but we are not in the business of taking money from businesses who cannot afford it. We recently reviewed a system that had been built for a small business, and our advice was to look at commercial products before spending any more on it.
Key takeaways
- Check the market before you build anything, because if a product already does the job it will almost certainly be cheaper.
- Building is worth considering when your business has processes no product covers, or when the software is what you sell.
- Owning software means paying to host, secure, back up, maintain and update it for as long as you use it.
- Software built on the cheap is rarely cheap in the end, whether it came from a favour, the lowest bidder or an AI app builder.
- If you do build, set a realistic budget and choose a partner you trust, even if they are not the cheapest.
Do you run your business on it, or are you selling it?
For most of the things a business needs to do, such as taking bookings, sending invoices or managing customers, software already exists. There are usually several products to choose from, and it is nearly always cheaper to try them out before talking to a developer.
If the software is simply the tool you use to run your day-to-day operations, an off-the-shelf product is very likely the better choice. It may not work exactly the way you do things now, but it is worth asking whether your process could change to fit the product, because that is often cheaper than changing the product to fit you.
If your business has its own processes that are not covered by any tool, and they are part of what makes you different, then building may be worth it.
The other case is when the software is part of what you are selling. If you have a good idea for a product, test the idea before you build it. That might simply start with some market research, talking to potential customers about what they use now and what they would pay for. A very simple proof of concept can also help, as long as everyone accepts it will be thrown away once it has done its job.
What owning software commits you to
When you own your software, you effectively become a software vendor in addition to running your main business. A subscription product comes with secure hosting, backups and data protection, all kept up to date by the vendor. If you build your own, those become your responsibility, along with fixing bugs, adding new features and modernising the technology as it ages. We cover what that adds up to in what it costs to maintain software after launch.
The difference in cost can be large. A system like the one we reviewed could easily cost £30,000 to £40,000 to build properly, which is in line with what most bespoke builds cost. A ready-made product for the same job will often cost somewhere between £20 and £100 per user each month. Even at the top of that range, five people would cost £6,000 a year, so it would take five to seven years to spend what the build cost, and that is before you have paid anything to maintain it.
These are not hard numbers, but a gap like that is typical, and it is what you should expect if you decide to build. Some people argue that owning becomes cheaper once you have enough users, because subscription costs grow with every person you add. There may be cases where that is true, but we doubt it happens often, because the costs of owning software carry on for as long as you use it. If you build, it should be because nothing on the market does what you need, and not to save money.
Software built on the cheap
There seems to be an opinion going round that building software is now cheap and easy, partly because of AI. There is also a temptation to hire developers at very low rates. On freelance marketplaces it can feel like a race to the bottom, with people asking for whole systems at prices nobody could build them for properly.
Building software is hard, and it requires discipline, commitment and a clear idea of what you are building. It also costs more than people might like to think.
The system we reviewed shows what can happen when software is not properly funded. It was not vibe coded, and the code itself was reasonably tidy. However it had been built cheaply, without rigorous development practices, so there was no source control, no automated tests and only a partial handover. The business was left with something that was not fit for purpose, and that they could not easily maintain or safely run.
AI has transformed software development, but these tools only make sense in the hands of experienced developers. They make skilled developers faster, and they do not remove the need for robust practices such as source control, automated deployments and testing.
People also often miss the ongoing costs of vibe-coded apps. The platforms that build them are often cheap or free to start with, but you are locking yourself into their platform, and you end up paying a monthly fee based on usage. These apps are not as cheap as they look, and moving one somewhere else can mean rebuilding parts of it.
If you already have something like this, get it reviewed before you spend more on it, and do not hold on to it just because of the money that has already gone into it.
If you do decide to build
Whichever way you go, start by writing down what you need the system to do. That is what any product would be assessed against, and what any build would be scoped from, and a paid discovery phase can help if you are not sure. Then check whether a product already does it, and how much it costs. A smaller system is cheaper to buy, build and secure, so it is worth asking which parts you really need.
If you still decide to build, set a budget and make sure it is realistic. Then find a partner you trust, who will tell you when something is not worth building, and do not just go with the cheapest option.
If you are weighing up this decision, tell us what you need and we will tell you whether we think it is worth building.
FAQs
Can you start with an off-the-shelf product and move to bespoke software later?
Yes, and it is often a sensible way to start. Using a product for a while teaches you what you actually need, which makes a later build much easier to scope and price.
Can a product be customised instead of building from scratch?
Often it can. Some products can be extended through an API, which lets a developer connect them to your other systems or add the piece that is missing. That is usually far cheaper than building the whole system, although you are still dependent on the vendor.
Who owns the data if you subscribe to a product?
You should, but check the contract says so before you sign up. Make sure you can export all of your data in a format you can use, so you can move to another product, or to a bespoke system, if you ever need to.