How these pages work
Every page follows the same sequence. It opens by confirming that the symptom is real and locating it: which pages, which devices, which sources and which step of the journey. A diagnosis table then sets out what you can see, the check that confirms or rules out each cause, the cause and the fix. Each cause gets a short section with the fix, a way to verify that the fix worked and the habit that stops the fault returning, followed by a routine that catches the whole family of causes before the next release or campaign.
The pages do not re-teach method. How a page should be built is in the anatomy of a landing page; how to locate a leak step by step is in the funnel analysis tutorial; how to plan, run and read a test is in the experimentation section. A troubleshooting page sends you to the relevant section once a cause is confirmed, and a fix that changes what visitors see should go through a controlled A/B TestControlled comparison of two variants. wherever the traffic allows one.
Which page to start from
Choose the row that matches what you can observe, not the cause you already suspect. The problem pages beneath this introduction are listed automatically, so each diagnosis appears here as soon as it is published; until then, the guide in the middle column holds the relevant checks.
| What you are seeing | Start with | Background |
|---|---|---|
| Conversion rate fell soon after a new site, template or relaunch went live | Conversion rate dropped after redesign | Website migration checklist |
| Conversion rate fell sharply on a particular day and nothing was redesigned | Conversion rate dropped suddenly | Funnel analysis |
| Conversion rate rose sharply and nobody can name a change that explains it | Conversion rate jumped overnight | Confirmation and thank-you pages |
| Visitors leave a landing page within seconds of arriving | Landing page bounce rate high | Message match |
| Many visitors start a form and few submit it | Form abandonment too high | Forms that convert |
| People say they sent the website form and nobody at the business received it | Website form submissions not arriving | Forms that convert |
| Phones convert far below desktops, month after month | Mobile conversion rate far below desktop | Speed and mobile |
| Free trials start, but few become paying customers once they end | Free trial users not converting to paid | Sign-up and trial-start flows |
| A test has run for weeks and still has no winner | A/B test not reaching significance | Sample size and significance |
| Test results look wrong: uneven arms, swings, a winner that did not hold | Common testing pitfalls | A/B testing guide |
Measurement, traffic or page: rule them out in that order
Every conversion rate (CVRShare of visits or clicks that convert.) has a numerator and a denominator, and both can move without a single visitor behaving differently. Check measurement first because it is the cheapest to confirm: compare the conversions in analytics with the orders, leads or bookings the business recorded for the same days. If the two disagree by more than usual, the problem is counting, and the analytics troubleshooting section owns the tag, consent and event faults. Check traffic second: break the period down by source, campaign and device. A new campaign, a change in an ad platform's targeting or a burst of automated visits changes who arrives, and a lower rate from a different crowd is not a page problem.
Only when the counting holds and the traffic mix is stable does the page come into question. Even then, the symptom usually points at a specific part of the journey. Fast exits from the first screen point at the Landing PageThe page a visitor arrives on from a campaign, built to turn that visit into one action. promise, speed or layout. Losses at a form point at fields, errors and doubts, the territory Nielsen Norman Group's form usability recommendations cover. Losses in a checkout point at costs, accounts and payment. When the clicks come from one ad platform and the fault lives only there, the platform's own hub, such as Google Ads troubleshooting or Meta Ads troubleshooting, is the better starting point.
When to run conversion research instead
A troubleshooting page answers one question about one symptom. When the rate has always been low rather than recently changed, when several symptoms appear together, or when nobody can say where in the journey visitors are lost, a research sprint is the better tool. The conversion research methods guide combines analytics, recordings, surveys and usability sessions to find where and why people leave; the research section turns those findings into hypotheses and a prioritised backlog; and the landing pages and funnel and checkout sections hold the fixes. The research sprint tutorial runs those methods on one funnel and one question until a backlog is ready, and the tutorials on session recordings and heatmaps, on-page polls and surveys, usability testing and expert review take one method at a time. The conversion rate optimisation pillar shows how research, testing, funnels and landing pages fit together.
