The commonest reason a landing page underperforms is not its design. It is that the advertisement promised one thing and the page is about another: a different price, a different product, a different audience, or a general page where a specific answer was expected. That mismatch costs more than any layout decision, it is free to fix, and it is routinely diagnosed as a design problem because redesigning is what people sell.
The page must answer the exact promise, in the first screen
If the advertisement said a price, the price is visible. If it named a city, the city is named. If it offered a specific service, that service is the headline rather than an item in a list. Somebody arriving has a question and roughly five seconds of patience. A general page invites them to find their own answer, which most of them will not do at the cost of a paid click.
One page, one action
A page with a form, a phone number, a chat widget, a newsletter sign-up and a link to the blog is asking the visitor to choose, and choosing is friction. Decide the one thing you want, make it the obvious action, and remove or subordinate everything else. The navigation is usually the first thing to go, and it is usually the change with the largest effect.
Ask for less than you want to know
Every field removes a proportion of the people who were going to fill it in, and most forms ask for information nobody needs until later. Name and one contact method is usually enough to start a conversation. If qualification matters, ask the qualifying question and drop the rest, rather than collecting a profile from somebody who has not decided to talk to you yet.
Speed on a phone is part of the conversion rate
Most paid traffic arrives on mobile, and a page that takes several seconds loses people before the headline renders. Image weight and third-party scripts are almost always the cause. Measure the page as a phone on a mediocre connection would experience it, rather than on office broadband, and treat the number as a conversion metric rather than a technical one.
Know whether you have the volume to test
If the page produces a handful of conversions a week, split testing cannot resolve anything smaller than a very large change, and an expert's judgement plus watching real sessions will find more. If it produces hundreds, testing is worth building. Any supplier who proposes a testing program without asking your conversion volume is selling a process.
Questions people ask about landing page experts
Why is our landing page not converting?
Most often because it does not answer the exact promise that brought people there. Check the advertisement and the page side by side before commissioning a redesign; the fix is frequently free.
Should a landing page have navigation?
Usually not. Every alternative action competes with the one you want, and removing the navigation is often the single largest improvement available on a paid landing page.
How many form fields is too many?
More than you need to start a conversation. Every field costs a proportion of submissions, and most forms collect information that could be asked later, from somebody who has already decided to talk to you.
Should we split test?
Only with enough conversions to resolve a result, which most pages do not have. Below that, expert review and watching real sessions find more than a test can prove.