A product audit is only worth it if you ship the fixes

“Get a second pair of senior eyes on it” is an easy yes. It’s low commitment, it flatters nobody’s ego, and it feels responsible. That is exactly why so many audits are wasted. The team gathers, nods along to fifty findings ranked in page order, files the deck in a drive, and ships next week’s roadmap unchanged.

The audit wasn’t wrong. It just never became anything. A finding that doesn’t change what you build is trivia, and trivia is cheap now, because a model can generate a hundred plausible ones before your coffee cools. What you are actually paying a senior person for is judgement about which few matter.

Not a list of everything imperfect. A useful audit is a prioritised, evidence-led fix list, ordered by impact and effort rather than by the order the screens happen to load. It comes from someone walking your critical paths the way a stressed first-timer would: signup, onboarding, the first transaction, marking exactly where friction and doubt creep in. That is the design-the-bad-day-first habit applied to a product that already exists.

The numbers back the method. Baymard’s checkout research puts the average large site’s achievable gain from better checkout design alone at around 35%. NN/g found the first study with just five users surfaces roughly 85% of the usability problems. The value was never in spotting issues. It is in knowing which three to fix first.

Red flags: it is priced by the number of pages reviewed; the auditor never asks for your analytics or a test account; findings are ranked by page order rather than by what they cost you; and there is no opinion in it. A good auditor tells you what they would do, not just what is technically suboptimal.

The deck-that-ends-at-the-deck is the biggest tell. If there is no walkthrough and no handoff, the findings die in the room. A good audit ends with your team knowing what to ship on Monday, and a quick win or two they can ship that afternoon.

Worth it: you have a live product with real traffic, and you can feel the funnel leaking but the analytics only tell you where people leave, not why. Or your team has read the interface so many times they can’t see it cold any more. Or you are about to pour a quarter into a build and want to aim before you fire.

Not worth it: you are pre-launch with no users. You need design, not a critique of something nobody has used. You already know the fixes and just need hands to make them. Or, honestly, you know you won’t act on it. An audit you won’t ship from is a paid opinion, and an expensive way to feel productive.

Ship the fixes. That is the whole test. At Lightyear, restructuring the help centre around search cut support contacts by roughly 15%, not because the findings were clever, but because they became changes in the product. If you take one thing from this: don’t buy an audit, buy the first three fixes and a plan for the rest.

If that is the shape of what you need, it is how I run them: a fixed-price product-design audit that ends in a prioritised fix list, not a lecture, with a €300 live teardown if you just want senior eyes on one flow this week.