A lean framework for diagnosing why a landing page is not converting

This is a sample entry that shows how a playbook looks. It will be replaced by a real one.

When a landing page is not converting, the usual response is a redesign. A redesign changes everything at once, which makes it hard to learn what was wrong. This framework asks five questions in order. Stop at the first one where the answer is no, fix that, and measure again before moving on. It grows out of the objection-led landing page experiment in the Lab and will be revised as that experiment reports.

Steps

1. Are the right people arriving? Break conversions down by traffic source. If one source converts and another does not, the problem may be the ad or the audience, not the page. Fix the traffic before touching the page.
2. Do they understand the offer in the first screen? Show the first screen to five people who match the buyer for five seconds, then ask what is being offered and for whom. If they cannot say, rewrite the headline and opening before anything else.
3. Do they believe it? Look for proof close to the main claim: reviews, examples, credentials, guarantees. Ask the people who handle enquiries which doubts come up most. If the page does not answer them, add the answers where the doubt would arise.
4. Is the next step clear and low-risk? Count the actions on the page. Check that the main one is visible without scrolling on a phone and says what happens next. A vague “Contact us” asks for more trust than a specific “Book a 15-minute call”.
5. Does the form or checkout let them finish? Fill it in yourself on a phone. Remove every field you do not need now, and check that errors are explained in plain words.

Common failure modes

• Starting at step four. Button colours and form fields are the easiest things to change, so they get changed first, while the real problem sits earlier.
• Judging on too little data. A page with a handful of visits a week can look broken or brilliant by chance. Wait for enough traffic, or treat the result as a hint, not a finding.
• Changing several things at once. If the headline, the proof and the form all change together, nobody will know which change mattered.
• Counting conversions, not good conversions. A page can convert more people and still produce fewer customers. Measure fit as well as volume.
• Skipping the conversations. Analytics show where people leave. Only enquiries, calls and interviews show why.