Before You Redesign: A Practical UX Audit Checklist

Start by writing the decision the redesign is meant to support. ‘Modernize the site’ is an output, not a decision. A useful decision might be: can a first-time operations lead understand our product category and request a demo without a sales call? Or: which parts of onboarding prevent a new workspace from reaching its first useful action? This language keeps the review anchored in a customer outcome instead of taste.

Then set a boundary. Decide which audience, pages, and moment in the journey are in scope. A homepage and a pricing page may share a visual system, but they answer different questions. A broad review can identify themes across the site; a focused review should still identify the one journey that needs to become clearer first.

Take a current-state inventory while the existing experience is still available. Record the live URLs, key screen states, current copy, CTA destinations, relevant mobile views, and any support or sales context your team already has. Screenshots are especially useful for preserving what a visitor actually sees; they make later discussions less dependent on memory or on a moving target.

Separate supplied context from evidence. A founder may know that prospects ask the same question on calls. A product manager may have noticed a recurring support theme. Those are valuable inputs, but they are not the same as a confirmed explanation for a page problem. Label them clearly so the new design does not turn a plausible story into an unquestioned requirement.

Review the path in the order a new visitor experiences it. At each step, use four lenses: comprehension, confidence, effort, and continuity. Comprehension asks whether the visitor can explain what is happening. Confidence asks whether the page gives an appropriate reason to believe the claim. Effort asks whether the next step feels proportionate. Continuity asks whether the message and context carry forward after a click.

This framework is more useful than a generic list of visual critiques because it identifies the kind of work required. A comprehension problem may need clearer positioning. A confidence problem may need a concrete example or constraint. An effort problem may need a shorter form or a more proportional CTA. A continuity problem may need the destination page to repeat the promise that earned the click.

A redesign should have a preservation list, not only a problem list. Identify pages, messages, or interaction patterns that are already understandable and useful. They may not be the most visually exciting parts of the experience, but they can contain hard-won clarity: a succinct workflow description, a plain-language explanation of pricing, or a CTA that accurately sets expectations.

For every proposed change, ask what customer understanding might be lost. This is particularly important when teams replace direct copy with abstract brand language, collapse helpful detail into a visual treatment, or move an important proof point lower to make a page feel cleaner. A simpler page is only better when it still answers the question that brought the visitor there.

Finish by grouping observations into a small number of findings. Each finding should state where it appears, the visitor question it affects, the evidence or context behind it, and the change you want to explore. Do not turn every note into a ticket. A list of twenty disconnected adjustments invites a fragmented redesign; three to five decision-relevant themes give the team something coherent to solve.

For high-priority findings, define how you will review the new version. That might be a fresh cross-functional walkthrough, a short customer conversation, or checking whether the revised page can answer the original visitor questions without additional explanation. The method will vary, but writing it down prevents the redesign from being judged only by whether it looks new.

You do not need a lengthy research program to make a better first redesign decision. Give a small group the same current-state materials, let each person complete the journey independently, and compare notes only after they have recorded what they saw. The difference between independent observations and shared assumptions is often where the most useful conversation begins.

If you need a clear record to bring into that conversation, VerdictUX offers a $29 Full UX Audit built around a saved preliminary review. You can keep reference uploads with the review, organize findings for discussion, and export a browser-based PDF. It is a practical starting point for a focused team review, not a replacement for customer research or specialist accessibility evaluation.