Build it, or buy it?
Most businesses should buy. This is how to work out whether you are one of them.
Off-the-shelf software is cheaper, faster and lower risk when it fits. Building is worth it when the gap between the software and how you work costs real money every month. The test is what the workaround is costing you.
Updated August 2026
Short answer
Buy if an existing tool covers roughly eighty percent and the rest is a nice-to-have. Build when the gap is a person re-typing data, a process taking a day it should not, or something you do differently that is why customers choose you.
Count the hours the workaround costs. That number decides it.
Most businesses should buy, and we say that as a company that builds custom software. Off-the-shelf is cheaper, faster and far less risky when it fits — and for standard processes it usually fits well enough.
The decision becomes arithmetic rather than opinion when you count what the gap costs. If someone spends a day a week working around the software — re-entering data, maintaining a spreadsheet beside it, doing manually what the tool will not do — that is roughly a fifth of a salary every year, permanently, and it grows as you hire.
Set that against a one-off build plus modest ongoing support and the comparison becomes clear in either direction. What is not a good reason to build is dislike of the interface, or a feature you would use occasionally, or wanting something that feels bespoke.
There is also a middle route people forget: buy the platform and build the one piece it cannot do, connected by an API. That is frequently the right answer and considerably cheaper than either extreme.
Side by side.
| Off-the-shelf | Custom build | |
|---|---|---|
| Upfront cost | Low — a subscription | Higher — a project |
| Cost after three years | Rises with users and tiers | Fixed, plus optional support |
| Time to running | Days | 2–4 months |
| Fit to how you work | You adapt to it | It fits your process |
| Your differentiator | Competitors run the same tool | Encoded in software only you have |
| Ownership | You rent it | You own it outright |
| If the vendor changes terms | You absorb it | Does not apply |
| Support | Their helpdesk | Whoever you choose |
Buy off-the-shelf when
- An existing tool covers most of what you need
- Your process is fairly standard for your industry
- You need something running in days
- Nobody in your business wants to own software
Build custom when
- A person re-enters data between systems every week
- Licence costs rise faster than your team
- The way you work is genuinely why customers choose you
- You have tried configuring a platform and it still does not fit
What people get wrong.
Comparing subscription to build cost
The comparison should be three to five years of subscription plus workaround hours against a build plus support. Over that period they land closer than most expect.
Building because the interface is disliked
An expensive way to solve a preference. If the software does the job and people simply find it ugly, that is rarely worth a project.
Missing the hybrid option
Keeping the platform and building only the piece it cannot handle is often the cheapest good answer, and it gets skipped because the question is framed as either-or.
Forgetting that bought software also needs maintaining
It is called a subscription and it rises with users and tiers. Custom software has support costs; bought software has them too, just less visibly.
Four questions that settle it.
If you cannot answer the first with a number, the answer is buy.
How many hours a week does the workaround cost?
Count them properly. If nobody can point at concrete hours, the gap is not yet large enough to justify building.
Does an existing tool cover roughly eighty percent?
If yes and the remainder is a nice-to-have, buy it. If the remaining twenty percent is where your business differs, that is the case for building.
Is the way you work genuinely why customers choose you?
If your process is the advantage, generic software makes you generic. If your process is standard, standard software is fine.
Could you buy the platform and build one piece?
Ask this before committing to a full build. It is often the cheapest good answer and it is frequently overlooked.
What people ask next.
Count the hours. If someone spends a day a week working around the software, that is roughly a fifth of a salary, every year, forever — and it usually grows. Set that against a one-off build cost and the answer becomes arithmetic rather than opinion. If nobody can point at concrete hours being lost, the gap is probably not big enough yet.
Yes, and it is often the sensible sequence. Buy something now, learn what actually matters by using it, and build later with real knowledge rather than assumptions. The one thing to protect is your data — make sure you can export it in a usable form before committing, because that is what makes moving possible.
A legitimate middle route and worth considering seriously. Platforms like Salesforce or Odoo can be configured a long way, and for many businesses that lands in the right place. The costs to weigh are the consultants, the per-user licences that grow with your team, and the fact that heavy configuration can be as hard to change later as custom code.
Not necessarily, and the comparison is often framed unfairly. Off-the-shelf has maintenance too — it is just called a subscription, and it rises. Custom software needs occasional updates and someone available when something breaks, which is a support arrangement rather than a licence. Over three to five years they frequently land closer than people expect.
Let's build what's next.
Tell us what you’re building. We’ll tell you how we’d help.