Free self-assessment

Release Risk Self-Check

Five dimensions. Twenty-five questions. About fifteen minutes. For teams shipping faster than their quality practices can keep up with.

How to score

Twenty-five statements, five per dimension. For each:

Yes.
True, consistently, and you could show someone the evidence inside ten minutes. Not “we intend to”, not “it’s on the roadmap”, not “in theory”.
Partly.
True some of the time. Or true for one team and not the others. Or true in principle but you couldn’t prove it today. Or, the common one, it used to be true and has quietly drifted.
No, or you don’t know.
Not knowing counts as a no. That isn’t a punishment, it’s the finding: if nobody can say whether a failing test blocks the release, then in practice it doesn’t reliably.

When you're between two answers, take the lower one. Self-assessment runs optimistic, and this is only worth the fifteen minutes if it finds something.

0 of 25
1 / 5Direction0 / 10
  1. 1.1 We can name our top three priorities this quarter, and they are not “everything”.

  2. 1.2 Every roadmap item has a written reason that refers to a user.

  3. 1.3 We decide what to build next from evidence (usage data, support volume, customer conversations) rather than the loudest voice.

  4. 1.4 We define what success looks like before we build, and we go back afterwards to check.

  5. 1.5 In the last six months we have cut or killed a planned feature because the evidence didn't support it.

2 / 5Architecture0 / 10
  1. 2.1 Someone could draw the system on a whiteboard in five minutes from memory, and other engineers would recognise it.

  2. 2.2 We have a system diagram that was updated in the last six months.

  3. 2.3 Two engineers, asked separately, would describe the main components the same way.

  4. 2.4 We can trace our most critical transaction end to end and name every service it touches.

  5. 2.5 We can explain why each major technology choice was made, and by whom.

3 / 5Testing strategy0 / 10
  1. 3.1 We could explain in two minutes what we test at each level, and why.

  2. 3.2 Our test mix was chosen deliberately. It is not simply what accumulated.

  3. 3.3 We know where our testing is weakest, and that is a conscious trade-off rather than a blind spot.

  4. 3.4 We know roughly how many bugs reached production last quarter, and whether that number is rising.

  5. 3.5 Our tests derive from stated requirements or acceptance criteria, not only from the code that was written.

4 / 5Test automation0 / 10
  1. 4.1 Automated tests run on every merge to the main branch, without anyone triggering them.

  2. 4.2 Pipeline feedback arrives in under fifteen minutes.

  3. 4.3 A failing test blocks the release, and that rule has not been quietly suspended.

  4. 4.4 We know which of our tests are flaky, and we fix them rather than re-running.

  5. 4.5 Someone would notice within a day if the test suite stopped running altogether.

5 / 5Lifecycle and release0 / 10
  1. 5.1 We could trace a feature shipped last month from requirement to code to test to deploy, without a gap.

  2. 5.2 Releasing to production is mostly automated and does not depend on one person's knowledge.

  3. 5.3 We can roll back a bad release in under fifteen minutes, and someone has actually done it in the last six months.

  4. 5.4 We can turn a feature off without deploying.

  5. 5.5 When production breaks outside working hours there is a named path that does not rely on individual memory.

Your result

– of 50

Answer all twenty-five statements to see your result.

Finish all twenty-five statements and we'll send you your score, plus the cheapest first move for your weakest dimensions.

FAQ

Questions we often get

How long does it take?

About fifteen minutes: 25 statements across five dimensions.

Is it free?

Yes. You get your score, what it means, and the cheapest first move for your weakest dimensions.

Who is it for?

Product teams shipping software in startups and mid-market companies, and the people who lead them.

What happens with my answers?

We send you your result and keep your contact details and answers for a while. One of us may email you personally about your result. If you tick the box for occasional emails, we also add you to our newsletter, which you can leave at any time. The privacy policy has the details.