Website Conversion Audit Checklist: 30 Points Before Buying Traffic
A 30-point website conversion audit checklist: data to pull first, offer, proof, forms, speed, mobile UX, trust and tracking, plus how to prioritize fixes.

A website conversion audit checks, in a fixed order, why visitors who already arrive do not become leads or buyers: first the data (where people drop off and on which devices), then the page itself (offer clarity, proof, CTA and forms, speed, mobile UX, trust) and finally the tracking that reports all of this. Run it before you raise the ad budget, because extra traffic sent to a leaking page mostly buys more of the same leak. The 30-point checklist below covers what we check on client sites, grouped by area, with a quick fix for each group.
When a conversion audit is worth doing
An audit pays off when the problem is on the site, not in the traffic. Typical signals:
- Traffic but few leads. Sessions grow or hold steady, form submissions and calls do not.
- Rising cost per acquisition. Ad costs per click are stable, but the cost per lead keeps climbing. Before blaming the campaign, check what happens after the click.
- Before a redesign. A new design built on guesses repeats the old mistakes in a new style. An audit tells the designer what to keep, what to remove and what to fix first.
- After a launch or migration. Broken forms, lost tracking and missing redirects show up here more often than anywhere else.
If the site has almost no traffic at all, an audit will not help much: there is too little behavior to analyze, and the bottleneck is acquisition.
Data first: what to pull before looking at the design
Opinions about a page are cheap; data tells you where to look. Before anyone comments on colors or copy, collect:
- GA4 funnel and key events. GA4 now calls conversions "key events". Check that form submissions, calls, bookings and purchases are marked as key events, then build a funnel exploration from landing page to form view to submission to see which step loses the most people.
- Sources by conversion rate. Compare key event rate by source and campaign. A channel with lots of sessions and almost no leads is either the wrong audience or a mismatch between the ad and the page.
- Devices. Split everything by mobile and desktop. A large gap in conversion rate between the two usually points to a mobile UX or form problem, not to the offer.
- Landing pages. Sort entry pages by sessions and key event rate. The pages with high traffic and low conversion are where fixes pay back first.
- Session recordings and heatmaps. Microsoft Clarity (free) and Hotjar (free and paid plans) both offer session recordings and click and scroll heatmaps. Clarity also flags rage clicks and dead clicks, which quickly surface broken or misleading elements.
- Form analytics. Track field-level behavior: which field people start, which one they abandon, which validation error they hit. Without it, a form is a black box with one number at the end.
The checklist: first screen, proof, CTA and forms
First screen and offer clarity
- 1. The headline says what you offer and for whom in plain words, not a slogan.
- 2. The first screen continues the promise of the ad or search result that brought the visitor.
- 3. One primary call to action is visible without scrolling on desktop and mobile.
- 4. The main benefit or outcome is stated, not only the features.
- 5. Nothing on the first screen competes with the CTA: no auto-rotating sliders, no pop-up in the first seconds.
Proof
- 6. Client logos or recognizable names sit near the first screen, if you have the right to show them.
- 7. Case studies show the task, the solution and the result in concrete terms.
- 8. Reviews are attributable: name, company or role, ideally a source such as Google reviews.
- 9. Proof sits next to the decision point (the form, the price), not only on a separate page.
CTA and forms
- 10. Every field is needed for the first contact; everything else is asked later.
- 11. The button text names the result ("Get a quote", "Book a call"), not "Submit".
- 12. Error messages appear next to the field, explain the problem and keep the entered data.
- 13. Fields use the right input types, so phones open the right keyboard: email, tel, numeric input for ZIP codes, plus autocomplete attributes.
- 14. After submission the visitor sees a clear next step: what happens, when, and who will contact them.
- 15. Submissions actually arrive: in the inbox, the CRM or the chat, tested end to end.
The checklist: speed, mobile UX, trust and analytics
Speed and Core Web Vitals
- 16. Field data for LCP, INP and CLS is in the "good" range on mobile for key landing pages.
- 17. The hero image is compressed and served in a modern format; nothing heavy blocks the first render.
- 18. Third-party scripts (chat widgets, trackers, embeds) are audited and the unused ones removed.
Speed is a deep topic of its own; if this group fails, a dedicated performance pass is usually cheaper than a redesign.
Mobile UX
- 19. Tap targets are large enough and not crowded together.
- 20. A sticky or repeated CTA is reachable without scrolling back to the top.
- 21. Phone numbers are tappable links, and the call path works.
- 22. No horizontal scroll, cut-off tables or popups that cannot be closed on a small screen.
Trust
- 23. Pricing is transparent: real prices or at least a "from" price and what affects the final number.
- 24. Several contact options are offered: form, phone, email, chat or calendar booking.
- 25. The company is real and visible: address or service area, team, legal details, privacy policy.
- 26. Content is current: no outdated dates, dead links or old offers.
Analytics correctness
- 27. Every lead action fires exactly one key event, without duplicates on page reload.
- 28. UTM parameters survive to the CRM, so leads can be traced to a campaign.
- 29. Internal traffic and test submissions are filtered out.
- 30. Calls and chats are counted, not only forms, otherwise the best channels look weak.
Audit summary table
| Area | What to check | Quick fix | Impact |
|---|---|---|---|
| First screen | Clear offer, one CTA, ad match | Rewrite the headline, remove the slider | High |
| Proof | Logos, cases, reviews near the form | Move the best proof next to the CTA | Medium to high |
| CTA and forms | Fields, errors, input types, delivery | Cut fields, fix validation and keyboards | High |
| Speed | Field Core Web Vitals on mobile | Compress the hero, drop unused scripts | Medium |
| Mobile UX | Tap targets, sticky CTA, tap to call | Add a sticky CTA, fix layout bugs | High on mobile-heavy traffic |
| Trust | Prices, contacts, company details | Add "from" prices and contact options | Medium to high |
| Analytics | Key events, UTMs, duplicates | Fix tags, pass UTMs to the CRM | Indirect, but decides every other call |
How to prioritize and test the fixes
An audit usually produces more findings than a team can ship at once. Two simple scoring models keep the list honest:
- ICE (Impact, Confidence, Ease), associated with Sean Ellis: rate each idea from 1 to 10 on how much it could move the metric, how sure you are, and how easy it is to ship.
- PIE (Potential, Importance, Ease), developed at WiderFunnel: rate how much a page can improve, how valuable its traffic is, and how hard the change is.
Both are subjective, and that is fine: their job is to force a ranking, not to predict a result. Broken things (a form that does not deliver, a missing key event) skip the scoring and get fixed first.
A/B testing basics. Test only what is uncertain; obvious bugs are fixed, not tested. Decide the sample size and test duration before launch with a calculator, run in full weeks, and do not stop the test the moment one variant looks ahead, because early leads often disappear. On low-traffic sites a valid test can take months, so ship clear fixes directly and compare periods carefully, keeping seasonality and campaign changes in mind.
Common false fixes
A common mistake we see is spending effort where it feels productive rather than where the data points:
- Changing button colors while the offer is unclear and the form has twelve fields.
- A full redesign without knowing which step loses people, which often just moves the problem.
- Adding more traffic to "get more data" when the leak is already visible in recordings.
- Chasing a perfect lab speed score while field metrics already pass.
- Adding pop-ups and chat widgets that interrupt the visitor and slow the page.
- Trusting conversion numbers without checking that events fire once and leads reach the CRM.
How to choose: our decision rule
Our rule is simple: fix tracking first, then the biggest leak on the highest-traffic page, then everything else. In practice:
- If you need a prioritized list of what blocks leads on your key pages, start with a conversion audit: $500, 3 days, covering the checklist above with concrete fixes.
- If the problems go deeper into navigation, user flows and interface logic across the product or site, a UX audit from $1,500 over 1 to 2 weeks gives a fuller picture.
- If the audit shows that the issues are structural (outdated design, slow platform, broken mobile layout), a website revamp in 2 weeks from $3,000 fixes them on the existing content and domain.
- If the site already converts reasonably and has enough traffic, landing page A/B testing from $1,500 per cycle is the next step.
Only after the leaks are closed does buying more traffic make sense: then every extra dollar lands on a page that can turn it into a lead.
Most often the page does not match what the visitor expected from the ad or search result, the offer is unclear, or the form is too long or broken on mobile. Tracking errors are another frequent cause: leads arrive but are not counted. Check the funnel by device and source first, then watch session recordings of the main landing pages.
A CRO audit focuses on the path to a lead or sale: offer, proof, CTA, forms and tracking on key pages. A UX audit looks wider at navigation, flows and interface usability across the whole site or product. Many projects start with a CRO audit and move to a UX audit when the problems turn out to be structural.
For behavior data it often is: it records sessions, builds click and scroll heatmaps and flags rage and dead clicks at no cost. It does not replace GA4 or the CRM for counting leads by source, so use it alongside correct key events and UTM tracking.
Do a full audit before major budget increases, before a redesign and after a migration or relaunch. Between those, a lighter review of key events, form delivery and top landing pages every quarter catches most regressions.
Fix clear problems directly: broken forms, missing proof, unreadable mobile layouts. Use A/B tests for uncertain choices such as a new offer angle or a different form length, and only when the page has enough traffic to finish the test in a reasonable time.



