Decision tool

A four-minute diagnostic for business leaders

Build, Buy
or Borrow?

Your business wants better outcomes — faster turnaround, fewer errors, more work done by the same team — and technology looks like a big part of the answer. But where do you start?

Build something yourself? Buy a product off the shelf? Bring in a partner to build it for you? Every one of those is right for some businesses and expensive for others. Getting the first choice wrong is how a year and a budget disappear.

This tool asks about your situation, not about technology. It gives you four decisions, explains what you gain and what you take on with each, and shows you the options your own answers have already ruled out.

Thanks to AI, almost anyone can build software now, and that possibility has both staff and suppliers emboldened by what suddenly looks possible.

But possibility is not a decision. What gets built, who builds it, and where it runs still have to be weighed against your business need, your regulatory position, your appetite for risk, and your own context. Enthusiasm is an expensive thing to mistake for judgement.

Use this to start your evaluation.

1
How it should be builtA quick prototype, a properly reviewed build, or full engineering rigour
2
Who should build itA specialist supplier, a partnership, or your own people
3
What it should be built onA finished product, a platform you configure, or a blank canvas
4
Where it should runYour own machines, a service someone else runs, or a cloud platform you operate

About four minutes · roughly 20 questions
Your answers stay in your browser unless you ask for the report
Sponsored by insdi

Why this isn't just a scoring quiz

Most assessments add your answers up and give you the average. That hides the answers that should stop a decision dead. A business with something genuinely valuable to protect, no developers of its own and a deadline will average out to "outsource all of it" — which is the one thing it must not do.

So this works in three layers. First, your answers are weighted and scored — how badly failure hurts counts for far more than how often the thing will change. Second, hard rules remove options that no score can buy back: if the data is regulated, a quick prototype is off the table no matter what else you say. Third, where two of your answers can't both be true, you're shown the conflict and asked to choose, rather than having it quietly averaged away.

Question 1

Your answers disagree