Skip to content
Strategy22 July 2026·6 min read

Most AI projects fail because they were the wrong one

The expensive mistake isn't picking the wrong model. It's automating something that was never the bottleneck.

We turn down a meaningful share of the AI work we're asked to do. Not because it's hard, but because it wouldn't have paid for itself, and finding that out after six months is far more expensive than finding it out in week one.

The pattern

A business hears it should be doing something with AI. Someone picks the most visible process rather than the most costly one. Six months later there's a working system nobody uses, because the thing it automated took two hours a week.

What actually predicts success

  • The process runs many times a day, not many times a month
  • It is currently done by a person, in a consistent way
  • There is a number attached to it — hours, rupees, lost leads
  • Someone in the business will own the output after go-live

The cheap version of finding out

A two-week audit costs a fraction of a build and tells you which processes clear that bar. Ours ends with a costed shortlist you own outright — you can take it to anyone, including someone other than us.

Want this applied to your business?

Tell us what your team does by hand and we'll tell you honestly whether it's worth automating.

Request a quote

Next step

Let's work out what's worth building.

A 30-minute call. We'll tell you what we'd do, roughly what it costs, and whether it's worth doing at all.