Currently in:

Hungarian flag

Budapest, Hungary

Local time:

November 18, 10:17
November 18, 10:17

What a product design audit should actually tell you

A useful product design audit finds the decisions creating friction, explains their impact, and gives your team a practical order for fixing them.

All articles

Written by

Portrait of Sebastian Bistran

Sebastian Bistran

Senior product designer and Framer specialist

Published

Reading time

 
Share
Team sticking notes on a glass wall during a design review

A product design audit is not a screenshot covered in red circles. It should help your team decide what to fix, why it matters, and what can wait.

The audit starts with the job

Before reviewing screens, define what the product is meant to help someone accomplish. A dashboard for a clinician has a different job from an onboarding flow for a patient. Without that context, an audit becomes a list of personal preferences.

The useful question is not whether a screen looks modern. It is whether the design helps the right person make the right decision with less uncertainty.

How a product design audit runs, step by step

Every product is different, but the shape of a good product design audit stays consistent. This is the sequence I follow:

  1. Agree the goals, audiences and journeys that matter most to the business

  2. Gather the evidence that already exists: analytics, support tickets, recordings and research

  3. Walk each critical journey as a real user would, on desktop and mobile

  4. Review the screens against usability heuristics and accessibility basics

  5. Group the findings into themes and rate them by impact, confidence and effort

  6. Play the results back to product, design and engineering together

The playback matters as much as the document. It is where the team agrees what the findings mean and turns them into decisions.

Person using a phone in front of a colourful painted wall

Five questions a product design audit should answer

Where are people hesitating?

Look for moments where the next action is unclear, labels require interpretation, or feedback arrives too late. Hesitation often appears in onboarding, forms, empty states, and tasks with several dependent steps.

What is unclear rather than ugly?

Visual polish matters, but clarity comes first. Weak hierarchy, competing actions, unfamiliar language, and missing status feedback can make a perfectly attractive interface difficult to use.

Which patterns are inconsistent?

The same action should not change its label, position, or behaviour from one screen to the next. An audit should find repeated components that have drifted and recommend a shared pattern the team can maintain.

What creates risk?

Some friction is inconvenient. Other friction can cause an incorrect submission, lost work, accessibility failure, or a decision made with incomplete information. Those findings deserve a different priority.

What should be fixed first?

A useful audit separates quick improvements from structural work. It should explain impact, confidence, and likely effort so the team can plan rather than receive a flat list of thirty urgent problems.

Accessibility belongs in every audit

Colour contrast, focus states, form labels, error messages and touch targets are not a separate compliance task. They are usability issues that affect everyone, and they are some of the cheapest problems to fix once they are named.

Checking them during the audit means the recommendations already respect accessibility, rather than becoming a second round of rework after the redesign ships.

Evidence beats opinion

A heuristic review is a good starting point, not the whole story. Combine it with the evidence you already have: analytics, support conversations, session recordings, user interviews, failed tasks, and questions repeatedly raised by the team.

When data is limited, say so. A clearly labelled hypothesis is more useful than an assumption presented as fact.

How to prioritise what the audit finds

Every finding gets three quick ratings: impact on the user or the business, confidence in the evidence behind it, and the effort needed to fix it. High impact, high confidence and low effort goes to the top of the list. High impact with low confidence becomes something to test before anyone builds it.

This keeps the conversation away from taste. Instead of debating whether a screen feels right, the team can see why one fix matters more than another and what it would take to ship it.

It also protects the roadmap. Quick wins can ship within a sprint, while structural changes get scoped properly instead of being squeezed in alongside feature work.

What the deliverable should include

  • The user or business outcome affected by each finding

  • Evidence showing where the issue came from

  • A severity and confidence rating

  • A practical recommendation, not just a criticism

  • Quick wins separated from deeper product changes

  • Examples or annotated flows that make the next step clear

The document should be easy for product, design, and engineering to discuss together. If it only makes sense to the designer who wrote it, it is not finished.

When an audit is enough

An audit is often enough when the product works but has accumulated friction, inconsistency, or design debt. You can improve the experience in focused releases without rebuilding everything.

A redesign becomes more likely when the underlying workflow no longer matches the product, the information architecture cannot support new use cases, or the interface is built from patterns the team cannot maintain.

Signs your product needs an audit

  • Support keeps answering the same questions about the same screens

  • New features take longer to design because every screen solves things differently

  • Activation or completion rates have stalled even though traffic is healthy

  • The team disagrees about what to fix first and nobody has evidence to settle it

  • A redesign is on the table, but nobody can say exactly what is broken

If two or more of these sound familiar, an audit will almost always pay for itself before a redesign does.

A health-tech example

IbisVision remote eye testing shown on a laptop and phone in a clinic waiting room

Remote eye testing has two very different audiences. Patients need reassurance, clear instructions, visible progress, and accessible language. Clinicians need fast review, useful hierarchy, and confidence in the data. The interface cannot simplify one side by hiding information the other side needs. That tension shaped the IbisVision case study.

A good audit makes tensions like this visible before a team jumps into new screens. It turns redesign from a visual exercise into a sequence of decisions.

How long does a product design audit take?

For a focused product with one or two core journeys, one to two weeks is usually enough: a few days to gather evidence and walk the flows, a few days to synthesise, and a session to play the findings back to the team.

Larger platforms with several audiences take longer, mainly because every audience needs its own journeys reviewed. If time is tight, audit the journey closest to revenue or risk first and treat the rest as a second phase.

Use the audit as a decision tool

The best outcome is not a longer report. It is a team that knows what to fix now, what to test next, and what can safely wait.

If your product has become harder to explain than it should be, send me the problem. I can help you find the friction before you commit to a full redesign.

Create a free website with Framer, the website builder loved by startups, designers and agencies.