When a campaign isn't converting, most businesses reach for the same fix: rebuild the landing page. Sometimes that's right. Just as often, the landing page is fine and the real problem is who you're showing the ad to in the first place. Guess wrong and you'll spend weeks and real money rebuilding a page that was never broken, while the actual leak keeps bleeding traffic.
The good news is you don't have to guess. GA4's landing page reports, your search terms report and session recordings will tell you, in plain numbers, which side of the problem you're actually dealing with. This post walks through how to read them, with a real-world example of each failure mode, so you fix the thing that's actually costing you conversions.
When a campaign isn't converting, most businesses reach for the same fix: rebuild the landing page. Sometimes that's right. Just as often, the landing page is fine and the real problem is who you're showing the ad to in the first place. Guess wrong and you'll spend weeks and real money rebuilding a page that was never broken, while the actual leak keeps bleeding traffic.
The good news is you don't have to guess. GA4's landing page reports, your search terms report and session recordings will tell you, in plain numbers, which side of the problem you're actually dealing with. This post walks through how to read them, with a real-world example of each failure mode, so you fix the thing that's actually costing you conversions.
Key Summary
- Low conversions with good-quality traffic (relevant search terms, the right audience, decent time on page) is a landing page problem.
- Low conversions with bad-quality traffic (sky-high bounce rate, irrelevant search terms, mismatched audience) is a targeting problem, and no landing page rebuild fixes that.
- Check GA4's landing page report and your search terms report before you touch either the campaign or the page.
- Session recordings show you what traffic quality metrics can't: exactly where real visitors get stuck or lose interest.
- Rebuilding a landing page when the problem is targeting wastes budget twice: once on the ads still reaching the wrong people, and once on a rebuild that changes nothing.
- Diagnosis takes twenty minutes. Guessing wrong costs weeks.
Why "It's the Landing Page" Is Everyone's First Guess
When conversions are low, the landing page is the easiest thing to blame. It's visible, it's yours, and it feels fixable. Swap the headline, change the button colour, add a testimonial, done. Targeting feels murkier and harder to control, so it gets ignored until the landing page rebuild fails too.
The opposite mistake happens just as often. Someone decides the audience is wrong, narrows the targeting, tweaks the interests or the keywords, and wonders why conversions barely move. Both guesses skip the same step: actually looking at what the traffic is doing before it hits the page.
This isn't really a landing page question or a targeting question. It's a conversion rate optimisation question, and CRO starts with diagnosis, not design. The fix depends entirely on where in the funnel things are breaking down, and that's a data question, not a gut-feel one.
When It's Actually the Landing Page
If the traffic itself looks healthy, the landing page is your problem. Healthy traffic means the search terms triggering your ads are relevant, the audience matches your targeting brief, and people are sticking around for a reasonable amount of time before they leave. They're just not converting once they get there.
Take a car servicing client running Google Ads for "logbook service [suburb]." The search terms report showed exactly the queries you'd want: location-specific, high-intent, no junk. Average engagement time on the landing page sat around 90 seconds, well above the bounce-and-run pattern of a mismatched audience. But the conversion rate was under 1%.
A session recording review showed the problem in under ten minutes: the booking form was buried below three scrolls of vehicle-brand logos and a badge carousel, and the phone number in the header wasn't clickable on mobile. People were reading the page, getting interested, then giving up trying to actually book. That's a landing page conversion rate problem. The fix was moving the form up and making the CTA impossible to miss, not touching the campaign at all.
When It's Actually the Targeting
If the traffic itself looks wrong before it even reaches the page, no amount of landing page optimisation will save it. The signs are different: a bounce rate well above what's normal for the industry, search terms that have nothing to do with the offer, or an audience whose demographics or interests don't line up with who actually buys.
A retail client running Meta Ads for a premium skincare range had a landing page that converted well for their organic and email traffic: north of 4%. But a new interest-based ad set was converting at 0.3%, with a bounce rate over 85% and an average session length under ten seconds. Nobody was reading anything, let alone considering a purchase.
The landing page hadn't changed. The audience had. The interest targeting was pulling in people drawn to budget beauty deals, not the premium buyer the page and the offer were built for. No amount of landing page design work fixes an audience mismatch. The fix here was rebuilding the targeting, not the page, and conversion rate recovered within a week once the right audience was back in front of the same page.
How to Actually Diagnose It Before You Spend Anything
Start with GA4's landing page report. Filter to the campaign or channel in question and look at engagement rate and average engagement time by landing page. Good engagement with low conversions points to the page. Poor engagement across the board points upstream, to the traffic itself.
Next, pull the search terms report (Google Ads) or the placement and audience breakdown (Meta Ads). This tells you exactly who's actually seeing your ads and what they searched or scrolled past to get there. Irrelevant terms or wildly mismatched interests are a targeting red flag, no matter how good your engagement numbers look elsewhere.
Finally, if you've got session recordings running, watch a handful from the underperforming segment. Numbers tell you something's wrong; recordings tell you exactly where. A form nobody can find, a CTA below the fold, a page that loads too slowly on mobile. These show up in five minutes of watching, not five hours of spreadsheet analysis.
A Quick Checklist Before You Rebuild Anything
- Is engagement rate and time on page for this traffic normal or low compared to your other campaigns?
- Are the search terms or audience segments actually relevant to the offer?
- Is bounce rate dramatically higher than your account average?
- Have you watched at least five session recordings from the underperforming segment?
- Does the page convert fine for a different, known-good traffic source (organic, email, direct)?
If traffic quality checks out and the page still doesn't convert, fix the page. If traffic quality is the problem, fix the targeting first. A landing page rebuild won't fix an audience that was never going to buy.
Frequently Asked Questions
What GA4 report should I check first for a conversion problem?
Start with the Landing Page report under Engagement, filtered by session default channel group or campaign. It shows engagement rate, average engagement time and conversions side by side, which is enough to point you in the right direction.
What counts as a "high" bounce rate?
It depends on the industry and traffic type, but anything well above your account's own average for similar campaigns is worth investigating. Compare the underperforming segment to a known-good one rather than chasing an industry benchmark number.
Can both the landing page and the targeting be wrong at the same time?
Yes, and it happens more than people expect. Fix the targeting first, since a landing page can't prove itself with the wrong audience arriving. Once relevant traffic is landing, re-test the page on its own merits.
Do I need session recording software to do this properly?
It's not essential, but it's the fastest way to confirm a landing page problem once GA4 and the search terms report have pointed you there. Without it, look at scroll depth and click data in GA4 as a rougher substitute.
How long should I test before deciding it's fixed?
Give any targeting or landing page change enough volume to be meaningful, not just a few days of data. As a rule of thumb, wait until you've got a comparable sample size to the period you diagnosed the problem from.
Wrapping Up
Landing pages and targeting fail in different, identifiable ways, and the data tells you which one you're looking at before you spend a cent fixing it. Good traffic that won't convert is a page problem. Bad traffic that never should have arrived is a targeting problem. Check the GA4 landing page report, the search terms report and a handful of session recordings before you rebuild anything. It takes twenty minutes and it means you fix the thing that's actually broken.
