How to Do a Basic UX Audit Yourself (Without Being a UX Expert)

What you can realistically assess yourself

Let's be honest about the limits first, because I think a false sense of confidence is worse than acknowledged uncertainty.

Without UX training, you can reliably evaluate: whether your navigation makes sense to someone who doesn't already know the site, whether your key pages communicate clearly what you want visitors to do, whether your CTAs are visible and specific, whether the experience on a mobile phone is genuinely usable or just technically responsive, and whether your forms ask for more than they need to.

What you'll likely miss: subtle trust signal problems, accessibility issues beyond the obvious ones, mobile edge cases that only surface on specific devices, and the pattern recognition that comes from reviewing hundreds of different sites. You'll also struggle to see problems caused by familiarity -- the things you've stopped noticing because you've seen them too many times.

That's fine. A DIY review isn't trying to replace expert evaluation. It's trying to find the easy wins and give you a baseline to work from.

The tools you'll need (all free)

Microsoft Clarity is the first tool I'd recommend to any founder doing their own UX review. It's genuinely free with no session cap -- unlike Hotjar's free tier, which limits you to 35 daily sessions. Clarity gives you heatmaps showing where people click and scroll, session recordings so you can watch real visitors move through your site, and rage click detection that flags places where people are repeatedly clicking something that isn't responding. Install it, let it run for a week, then watch five session recordings on mobile. You'll see things that no amount of traffic data would have shown you.

Google PageSpeed Insights takes your URL and returns a Core Web Vitals score in about 30 seconds. Run it on mobile first. That's where most sites struggle, and where most of the traffic is. Anything below 50 on mobile is worth investigating seriously.

GA4 is useful if you have it set up and have enough traffic to give the data meaning. The device breakdown report will show you whether mobile converts significantly below desktop, which is one of the clearest signals of a mobile UX problem. If you haven't set up conversion tracking yet, that's the first thing to fix before anything else -- you can't diagnose what you're not measuring.

If you want a rundown of the fuller free toolkit, including Search Console and an accessibility checker, this article on free UX audit tools covers each one in more depth.

The five questions to ask about any page

These are the questions I return to on every page I review. They're not exhaustive, but they're a reliable starting framework.

What is the first thing a visitor sees, and does it immediately tell them what the page is for? This sounds obvious. It's remarkable how often the answer is no. Hero images that look polished but don't communicate anything. Headlines that describe the company rather than what it does for the visitor. The first thing someone sees should answer "am I in the right place?" without any further scrolling required.

Is there a clear next step, and is it immediately obvious? Every page should have one primary action. Not three options at similar visual weight -- one clear primary action. Find it on your page. Now cover up everything else. Is it still obvious? If not, it probably isn't obvious to a first-time visitor either.

Does this page actually work on a phone? Not "does the layout resize" -- does it work. Open it on your actual phone, not a simulated mobile view in your browser. Tap the main CTA. Try to complete the form. Can you read the text without zooming? Are the buttons large enough to tap reliably with your thumb? This takes two minutes and regularly surfaces problems that have been there for years.

How many fields does the form on this page have? Count them. Then ask which ones you actually need to complete the intended action. Any field you can't justify is friction with no upside. Baymard Institute's checkout research found the average ecommerce checkout has around 11 form fields. Most stores need about six or seven.

Would a stranger trust this page? Fresh eyes are genuinely useful here. Ask someone who doesn't know your business to look at the page and tell you what they think the company does and whether they'd buy from it or sign up. If they hesitate, that hesitation is data.

How to structure your DIY review

Go through your three most important pages -- usually the homepage, your main product or service page, and your checkout or sign-up page -- and work through those five questions for each one.

Write down every issue you spot, however small. Then go back through the list and mark anything that sits directly between a visitor and a conversion. Those are your priorities. Everything else can wait.

Don't try to fix everything at once. Pick the one or two changes most likely to make a material difference, make those first, then measure what changes. This is more useful than a big redesign based on a list of 30 issues you've ranked subjectively.

What you'll probably miss

I want to be direct about this, because a lot of DIY audit advice oversells what someone without UX training can reliably find.

Design blindness is the biggest constraint. You've seen your own site too many times to see it clearly. The thing that's confusing to a first-time visitor often feels completely intuitive to you because you know what it means. This is why even a brief outside perspective -- a colleague who doesn't know the site, a friend, or a structured expert review -- tends to surface things a self-review won't.

Accessibility problems are often invisible without specific tools and knowledge. Colour contrast issues, keyboard navigation problems, missing alt text, screen reader compatibility -- these affect a real proportion of your visitors and don't show up in session recordings.

Mobile edge cases are notoriously hard to catch without testing on multiple actual devices. What works on your iPhone may not work on a mid-range Android phone, which is where a significant chunk of your traffic is likely arriving from.

When to move from DIY to a proper audit

A DIY review is a good starting point, and there are clear signals that it's time to get outside help.

If you've made changes based on what your self-review found and the conversion rate still hasn't moved, the problem is likely something you're not seeing. If your traffic is reasonable but conversions are persistently low and you can't identify why, that's a diagnostic gap that benefits from a fresh perspective. And if you're about to spend meaningful budget on a redesign or a paid traffic campaign, a structured audit first is almost always better value than finding out afterwards that the experience wasn't ready for it.

"The most common thing I hear after a DIY audit is 'I fixed what I could see but I knew there was something else I was missing.' That's usually right. The things that are hardest to find from the inside are often the things doing the most damage."

Alice, No-Friction Lab

The No-Friction Lab UX Audit Toolkit is the structured version of everything in this article. It includes a 10-point heuristic audit checklist, an effort vs impact prioritisation matrix, 20 AI prompts to accelerate the process, a revenue forecast worksheet, and a UX foundations guide -- all in a Notion template designed for founders and small teams who want to go deeper than a basic self-review without committing to a full professional audit.

If you'd rather hand it over entirely, professional audits start at £99: Ecommerce UX Audit, SaaS UX Audit, and UX and CRO Audit.