60-second check

A 60-second check: can this actually be AI?

Eight questions. Pick the answer closest to the truth. You'll see a verdict at the end. Everything runs in your browser and nothing gets sent anywhere.

Why this exists

Most 'AI ideas' fail a basic test. Find out now, not later.

When we run assessments, the same handful of issues keep turning up. The goal isn't specific enough to measure. The process changes too often for a model to learn it. The data exists but nobody trusts it. The work is rare enough that a person can just do it. The exceptions are the job. The value doesn't justify the build. Privacy or governance kills it in month three.

None of those make an idea stupid. They just make it wrong for AI, at least right now. Some of them mean you should be looking at plain automation instead. This check walks you through the eight tests we see most often, and tells you which bucket your idea falls into.

The check

Eight questions.

If you're not sure on any of them, pick 'not sure'. Honest is more useful than optimistic.

Your idea — in your words

Optional. Helps you stay focused on one specific thing while you answer.

Question 1 of 8

Is the goal clear?

Can you describe what success looks like in one sentence, without the word 'better'?

Question 2 of 8

Is the process stable and well understood?

If the steps change every few weeks, a model can't learn what the job is.

Question 3 of 8

Is the data available and in good shape?

Having it isn't enough if nobody inside the business trusts it.

Question 4 of 8

Are the outcomes measurable?

You need a way to tell, at the end, whether it did the job well.

Question 5 of 8

Is the work repetitive or rules-based?

Often enough to be worth automating, but not so rigid that a script would do.

Question 6 of 8

Are exceptions manageable?

If the job is mostly judgement calls, a model will get most of them wrong.

Question 7 of 8

Is the value worth the effort?

Counting both the build and the ongoing care of it.

Question 8 of 8

Are ethics, privacy, and governance properly considered?

If the data involves people, 'considered' has to mean their comfort too.

Nothing leaves your browser. We don't see your answers.

What happens with your answers

Nothing. The check runs entirely in your browser. We don't see your answers, we don't store them, and there's nothing to capture your email with. If the verdict suggests your idea is worth a closer look, there's a link to our free written assessment. If it doesn't, close the tab with our blessing.

If the check says 'maybe', the assessment says why.

The free assessment is the longer, written version of the same honesty.

Get a free assessment Book a 20-minute call