When your landing page is not converting, the usual reaction is to rewrite the headline, change the button colour or commission a new design. Often that fixes the wrong thing, because the cause sits where the page can’t reach it: a form that stopped sending, tracking that never fired, or an ad that attracts people who were never going to buy.
So diagnose before you redesign. Rule out measurement and traffic first, then walk the page in the order a visitor meets it, from the first screen to the reply after the form. Each cause below comes with a symptom, a check you can run with GA4, your ad account and a phone, and a fix.
Landing page not converting? The short answer
A landing page usually underperforms for one of three reasons. Check them in this order:
- Measurement. It is converting, but enquiries aren’t being delivered or counted.
- Traffic. The wrong people are arriving, so no page would convert them.
- The page. The right people hit a problem before they act: slow loading, a desktop-first layout, a mismatch with the ad, too many exits, missing proof, an unclear next step, a long or broken form, or slow follow-up.
Changing the page can’t fix the first two, so they come first.
Start with the data: is it the page at all?
A low landing page conversion rate can be real, or a counting error.
1. Check the enquiries are really missing
Symptom. Analytics shows few conversions, but nobody has checked that figure against real enquiries.
Check. On your phone, using mobile data, send a real enquiry. Did the email reach the right inbox, not spam or a former colleague? Did the key event appear in GA4’s Realtime report? Is the conversion action in your ad account still recording, or flagged as inactive? Then compare last month’s key events with the enquiries your team received, including calls and messaging-app chats.
Fix. Repair delivery first, then track every route to an enquiry: form submissions, phone taps and messaging clicks. If most people call, a page that looked dead may be working. How to track website enquiries in GA4 covers the set-up. Visitors who decline analytics cookies may go uncounted, so expect a gap, but both figures should tell the same story.
2. Check the right people are arriving
Symptom. Plenty of clicks, very short visits, and off-target enquiries from job seekers, people outside your area or people wanting something free.
Check. Read the search terms report in Google Ads, since broad match can pull in searches you never intended. Check the location setting too: “Presence or interest” also reaches people who’ve only shown interest in your area. On social platforms, check audience and placement breakdowns. In GA4, compare engagement by campaign. If one engages and another doesn’t on the same page, the traffic is the problem.
Fix. Add negative keywords, tighten match types, target “Presence” if you only serve people in your area, and let the ad qualify visitors: naming who it’s for, or a starting price, may cost clicks but brings more people who could buy.
3. Find the stage where visitors leave
If both check out, find where people drop off before guessing why. GA4’s enhanced measurement logs a scroll only at 90% page depth, and its automatic form events need testing on your own form before you trust them. A free recording and heatmap tool such as Microsoft Clarity fills the gaps; it masks form entries by default, but should also follow your cookie consent set-up.
| What the data shows | Where to look first |
|---|---|
| Most visitors leave within seconds | Stage 1: speed, mobile layout, message match, headline |
| Visitors scroll, but few click the call to action | Stage 2: exits, proof, the next step |
| Many start the form, few finish it | Stage 3: the form |
| Enquiries arrive but rarely become customers | Traffic quality and follow-up |
Stage 1: the first screen
Every visitor passes through this stage, so landing page mistakes here cost the most.
It loads slowly on mobile data
Symptom. Short visits concentrated on phones, and far fewer GA4 sessions than ad clicks. Consent choices explain some of that gap; a large one can mean people leave before the page loads.
Check. Run the landing page through PageSpeed Insights and read the mobile results. Google’s Core Web Vitals rate a Largest Contentful Paint of 2.5 seconds or less as good. Test the full address from the ad too: some page caches skip addresses with query strings, so the ?gclid= version ad visitors land on can be slower.
Fix. Usual culprits are an oversized hero image or video, a slider, a stack of pixels and chat widgets, and a cache that doesn’t ignore tracking parameters. How page speed affects conversions shows how to measure the cost in enquiries; why is my website slow? helps find the cause.
The layout was designed for desktop
Symptom. Phones convert far worse than desktops on the same campaign (compare key events by device category in GA4). Some gap is normal if people research on a phone and enquire from a desk.
Check. On a real phone, can you see the headline, offer and button without scrolling? Look for a tall image above the headline, text over a busy photo, a pop-up or cookie banner covering the button, and a phone number you can’t tap to call.
Fix. Design the phone version first. Move the visual below the button on small screens, add a sticky call to action, and check each language version separately, since translations can run longer and wrap differently.
It doesn’t match the ad
Symptom. One ad group loses visitors much faster than others, or Google Ads rates landing page experience, one of the three parts of Quality Score, “below average”.
Check. Put the ad and page side by side. Does the headline keep the ad’s promise? Are the price, product and audience on the first screen, in the ad’s language? An ad for “fixed-price office fit-out quotes” that lands on “Workspaces that inspire” loses people who were ready to ask.
Fix. Give each distinct promise its own page, or at least its own headline. If the ad still points at your homepage, see landing page vs homepage vs service page.
The headline is vague
Symptom. Visitors don’t scroll, and those who enquire ask what you actually offer.
Check. Show the first screen to an outsider for five seconds, then ask what’s on offer, who it’s for and what they’d do next. If they can’t answer all three, neither can your visitors.
Fix. State the outcome and the audience, then add specifics. “Solutions for growing businesses” says nothing. “Commercial kitchen fit-outs, designed and installed under one contract” tells the right visitor they’ve arrived. How to write landing page copy covers the research behind it.
Stage 2: the scroll
Too many goals and exits
Symptom. Visitors click away rather than down: to the menu, the logo, other services or social icons.
Check. Count the links that don’t lead to your one action, then use a GA4 path exploration or a click heatmap to see where people go.
Fix. Give the page one goal. Remove the main navigation, drop competing offers such as a newsletter sign-up, and keep only alternative routes to the same outcome, like calling instead of using the form. Landing page structure sets out the sections a focused page needs.
Missing or vague proof
Symptom. Engaged visitors reach the end, scroll back up to re-read the claims, then leave.
Check. List every claim on the page, such as “fast”, “experienced” or “trusted by industry leaders”, and beside each the evidence a stranger could check. The blanks are your gaps.
Fix. Swap adjectives for specifics: named testimonials (with permission), client logos, real project photos, certifications, guarantees and figures you can back up. “Experienced installers” says little; photos of finished kitchens with the client named say a lot. Put proof beside the claim it supports, and near the form.
No clear next step
Symptom. Deep scrolling, but few clicks on the call to action.
Check. Read the button aloud: “Submit” and “Learn more” don’t say what happens next. Does it appear only in the hero? Does the ask fit the traffic? A cold visitor from a social ad may not be ready to “Book a meeting”.
Fix. Use a verb plus an outcome (“Get a fixed-price quote”), place it after the proof and at the end, and say what follows the click. For a big ask, offer a smaller first step, like a price guide or a short call. See call to action best practices.
Stage 3: the form and what happens after it
The form asks too much
Symptom. Plenty of form starts, far fewer submissions.
Check. For each field, ask whether you need the answer to send a useful first reply. Budget, company size, “How did you hear about us?” and a required phone number are the usual extras.
Fix. Cut to what the first reply needs and ask the rest in conversation. If you must qualify, a multi-step form that opens with the easiest questions can feel shorter. Web form design goes into detail.
The form fails for some visitors
Symptom. Recordings show repeated taps on the button, or someone says “the form didn’t work”.
Check. Fill it in on an iPhone and an Android phone, using autofill. Try a name with accents or in a non-Latin script, and a phone number with a country code and spaces. Trigger an error: is the message specific, and does the form keep what you typed? Is a CAPTCHA or spam filter rejecting real people?
Fix. Loosen validation to accept real-world input, make error messages explain the fix, and retest after every plugin, theme or form update.
Follow-up is slow
Symptom. Enquiries arrive but rarely become conversations, and sales says “the leads are poor”.
Check. For your last 20 enquiries, compare arrival time with the first human reply. Then send a test enquiry on a Friday evening.
Fix. Send an instant confirmation saying who will reply and by when, route each enquiry to someone who can answer it, and promise only reply times you can keep. Where your ad platform supports it, import which leads became customers so campaigns optimise for sales, not form fills. Website lead generation covers routing and follow-up.
Fix one leak at a time
- Measurement first. Until every enquiry is counted, you can’t judge any other change.
- Then the earliest leak. A clearer first screen helps everyone; a better form helps only those who reach it.
- One dated change at a time. Keep ad settings steady and compare equal periods: seasonality and budget shifts can mimic a page effect.
- Test only when you have the volume. A/B tests need more conversions than many lead-generation pages get. A/B testing landing pages explains how to check, and what to do instead.
The conversion rate optimisation guide covers the wider process.
Landing page audit checklist
What to do next
Expect more than one leak, and work through the checks from the top. If the problems sit in the page itself, such as a desktop-first layout, competing goals or a message that doesn’t match your ads, rebuilding around one goal is often quicker than patching. Our landing page design service builds exactly that: one focused page for a campaign, launch or new service, in your brand and fast on phones.
Not sure whether the page, the traffic or the tracking is at fault? A free website audit looks at your page’s speed, mobile experience and how easily visitors can get in touch, and tells you where to start.