UX Audit Checklist: What to Check and How to Prioritise

A UX audit is a structured review of your website or app that finds where real people get stuck or give up. The short version of how to run one: pick the journey that makes you money, walk it yourself on a phone and a laptop, check it against the UX audit checklist below, then rank what you find by impact and effort.

Before you start

Most audits fail before anyone opens the site. They fail because the scope is "the whole website" and the output is a 90-slide deck nobody reads.

So narrow it down first.

• Write down the goal. One sentence. "More demo bookings from the pricing page." "Fewer abandoned checkouts on mobile." If you can't say what better looks like, you can't tell a real problem from a matter of taste.
• Pick one or two journeys that make money. For a SaaS product that's usually landing page to sign-up, then sign-up to first useful moment. For a shop, product page to paid order. For a services firm, any page to a sent enquiry. Audit those properly before you touch the About page.
• Pull whatever data you already have. Analytics funnels show where people leave. Session recordings and heatmaps hint at why. Support tickets and sales call notes are underrated: they're your users telling you, in their own words, what confused them.
• Decide who's auditing. Fresh eyes matter. The person who designed the checkout is the worst person to audit it, because they know where everything is.

No analytics? No recordings? You can still do a useful audit. You'll just be leaning harder on judgement and on watching a few real people try the journey (five or so is usually enough to surface the worst problems, a point Nielsen Norman Group has made for years).

The UX audit checklist

Each group below is a set of yes/no questions. Answer them honestly while you walk your chosen journey. A "no" isn't automatically a fire; it's a line on your list to rank later.

Underneath most of these checks sit Jakob Nielsen's 10 usability heuristics from 1994. They've aged well. If you want one framework to keep open in another tab while you audit, use that one.

First impression and clarity

• Can a first-time visitor tell what you do, and who it's for, within a few seconds of landing?
• Is there one obvious next step on the page?
• Does the headline describe an outcome the visitor cares about, in words they'd use?
• Does the page look like it was made by a business that's still around?

Failure usually looks like a clever headline that says nothing ("Reimagine what's possible") above three equally loud buttons. Say you run a booking app for physiotherapists. If a clinic owner lands and can't tell within a glance that you handle bookings and reminders, they bounce, and your analytics just call it a high bounce rate.

Navigation and findability

This is the group people skim. Don't. It's where a lot of quiet damage happens.

• Do the menu labels use your customers' words rather than your internal team names?
• Can someone find your three or four most-asked-about pages in two clicks or fewer?
• If you have a listing page (products, articles, jobs, case studies), is there a search, and does it return sensible results for typos and plurals?
• Do filters and sort options match how people actually choose? (Price, size, location, date: whatever the real decision is.)
• When a search or filter returns nothing, does the page suggest what to do next?

The classic failure: a working search box that only matches exact titles. Someone types "invoices" and your article is called "Invoicing guide", so they get zero results and assume you don't cover it. The control works. Findability doesn't. Test it by searching for things the way a stranger would, misspellings included.

Forms and checkout

If you only have time for one group, make it this one. It sits right on top of the money.

• Is every field you ask for genuinely needed at this step?
• Are labels visible above or beside fields, not only as placeholder text that vanishes when you type?
• Do errors appear next to the field, in plain language, and say how to fix the problem?
• Does the form keep what the person typed when something goes wrong?
• Can someone check out as a guest, or at least see the full cost before being asked to create an account?
• Are delivery costs, taxes and fees shown before the final step?
• Does the confirmation screen say what happens next, and when?
• Does every custom control (date pickers, dropdowns, toggles) work with a keyboard and on a small touchscreen?

Failure looks like this. A user fills in a twelve-field enquiry form, hits submit, and the page reloads with a red banner at the top saying "Please correct the errors below." The form is now empty. They don't fill it in twice.

Content and copy

• Can someone scan the page using only the headings and get the gist?
• Are button labels specific ("Book a call", "Download the guide") rather than "Submit" or "Click here"?
• Is the jargon gone, or at least explained the first time it appears?
• Do you answer the obvious objections (price range, timelines, what's included, who it suits) on the page where people have them?
• Is the copy consistent? Same thing, same name, everywhere?

Failure usually means a services page that describes your process in loving detail but never says what the client walks away with. Or a pricing page that hides the one answer everyone wants behind "Contact us." If you sell to busy people, assume they read the headings and the buttons. Nothing else.

Mobile

• Does the key journey work start to finish on a mid-range phone, on a mobile connection?
• Are tap targets large enough and spaced apart so thumbs don't hit the wrong thing?
• Does the right keyboard appear for each field (numbers for phone, email layout for email)?
• Do pop-ups and cookie banners leave enough screen to actually use the page?
• Is anything important hidden on mobile that exists on desktop?

Do this on a real device. Not a resized browser window. A very common failure is a sticky header, a chat bubble and a cookie banner that together cover a third of the screen, with the "Continue" button tucked underneath one of them.

Speed

• Does the main content appear quickly, roughly within Google's "good" Largest Contentful Paint threshold of 2.5 seconds?
• Does the page respond quickly to taps and clicks (Interaction to Next Paint under 200 milliseconds)?
• Does the layout stay still while loading, or do things jump around (Cumulative Layout Shift under 0.1)?
• Are images sized for the screen they're shown on?

Those three thresholds are Google's Core Web Vitals, and you can check them free in PageSpeed Insights. Failure often looks like a hero video that takes ages to load on mobile, or a "Buy" button that shifts down just as someone taps it because an ad or banner loaded above it. Speed fixes tend to belong to developers, so if this group fills up with "no", it's usually a web development job rather than a design one. Speed also affects search visibility, which is where it overlaps with SEO.

Accessibility

• Does body text meet the WCAG contrast ratio of at least 4.5:1 against its background?
• Can you complete the whole key journey using only a keyboard, with a visible focus indicator?
• Do images that carry meaning have useful alt text?
• Do form fields have real labels that a screen reader announces?
• Is colour never the only way information is shown (for example, errors marked only in red)?
• Does the page still work when text is zoomed to 200%?

A typical failure is pale grey text on white that looks elegant in a design file and is hard to read on a phone in daylight. Another is a custom dropdown that can't be opened without a mouse. Accessibility deserves more than six bullets, and we've written a longer guide on accessibility in UX design if you want to go further.

Trust signals

• Can people find who you are and how to reach a human?
• Are reviews, logos or testimonials real, recent and specific?
• Is your privacy policy easy to find, and does the sign-up form say what you'll do with someone's email?
• Do security cues appear where money or personal data changes hands?
• If your product uses AI, is it clear what the AI does and how to correct it?

Failure looks like a footer with no address and a testimonial from "John D., CEO." For AI features in particular, trust has its own set of design problems; we cover those in designing AI features users trust.

How long does a UX audit take?

It depends on scope. A focused audit of one journey, done by someone experienced, is a matter of days, not weeks. Add user testing, several journeys, or both a website and an app, and it grows from there. Be wary of any audit that takes longer than the fixes it recommends.

Can you do a UX audit yourself?

Yes, and you should do a rough pass yourself even if you plan to hire someone. Walk the journey on your phone. Answer the questions above. You'll catch the obvious problems for free. What an outside designer adds is distance (they don't know your product, which is the point) and pattern recognition from having seen the same failures many times. If your team is still working out who owns what in design, our piece on UI vs UX design is a useful primer.

How to prioritise your UX audit findings

An honest audit of a real product turns up dozens of issues. You won't fix them all, and you shouldn't try.

Score each finding on two things.

1. Impact. How many people hit this, and how close is it to money? A broken checkout field on mobile is high impact. An inconsistent icon on the settings page is not.
2. Effort. Rough guess, from your designer and developer together: a few hours, or a few weeks?

Then sort into four buckets.

• High impact, low effort. Do these this week. Rewriting a vague button, fixing an error message, removing an unneeded form field.
• High impact, high effort. Plan these properly. Restructuring navigation, rebuilding checkout. These usually need design work and testing first.
• Low impact, low effort. Batch them and fix them when someone's already in that part of the code.
• Low impact, high effort. Write them down and leave them. Seriously.

A UX review that ends in a deck is a report. One that ends in a ranked list of tickets with owners is a fix list. Aim for the second.

Turn each finding into a ticket that holds four things: what's wrong, where it happens (a screenshot or recording helps), why it matters for the goal you wrote down at the start, and what "fixed" looks like. Then measure the same funnel again after the changes ship, so you know whether you fixed the problem or just moved it.

If the same kind of issue keeps showing up across many screens (inconsistent buttons, five shades of grey, forms that each behave differently, spacing that changes from page to page), that's a sign the fix isn't page by page. It's a shared set of components, which is what a design system for startups is for. And if you're using AI tools to speed up parts of the audit, such as summarising session recordings, our notes on AI in UX design cover where that helps and where it doesn't. An audit also fits neatly into a wider UX design process, as the "find out what's broken" step before redesign.

---

If you'd rather have someone else walk your key journey with fresh eyes, that's part of what our UI/UX design team does. Or, if you already know what needs fixing, you can tell us about your project and we'll take it from there.