The short answer
Buy off-the-shelf unless the specific way you work is genuinely part of what makes your business good. Most of the time, buying is the right answer, and we will tell you so rather than take the work.
Compare over three years, not one
This is where most decisions go wrong. A one-off build price gets compared to a monthly subscription, and the subscription looks free.
| Off-the-shelf, 20 users | Custom build | |
|---|---|---|
| Year one | £30 per user per month = £7,200 | £2,999 to £5,999, plus hosting |
| Year two, 30 users | £10,800 | Hosting plus any changes |
| Year three, 40 users | £14,400 | Hosting plus any changes |
| Three-year total | About £32,400 | Build cost plus running costs |
| What you own at the end | Nothing | Everything |
| If you want to leave | Export what they allow | It is already yours |
The comparison flips on headcount growth. If your team is stable, off-the-shelf usually wins. If it is growing, per-seat pricing works against you every year.
The four questions that settle it
1. Does an existing tool do 80% of it? If yes, buy it and work around the 20%, unless the 20% is the entire point.
2. Is the way you do this a competitive advantage? If your process is genuinely better than your competitors', generic software will make you work like them. If your process is ordinary, generic software is fine and probably better than what you have.
3. What is your seat count doing? Growing fast makes per-seat pricing expensive fast.
4. What happens if the vendor changes the deal? Price rise, feature removal, acquisition, shutdown. If the answer is "we would be in serious trouble", that is a real risk with a real cost.
When we tell people to buy instead
Common questions
Is custom software always more expensive?
Over one year, usually. Over three years with a growing team, frequently not, because per-seat pricing scales with your headcount while a one-off build does not.
What is the real risk of off-the-shelf?
Two things: you reshape your process to fit the tool, and you cannot leave. Price rises, feature removals, and acquisitions all happen to software you depend on and do not control.
How do I compare them fairly?
Over three years, including per-seat growth on one side and hosting plus maintenance on the other. Comparing a one-off build price to one month of a subscription is how people talk themselves into the wrong answer.