The phrase landing page expert covers three different jobs that get sold interchangeably. One is a designer who makes a page look considered. One is a copywriter who rewrites the argument. One is a conversion specialist who runs a testing programme and can tell you, with evidence, which change moved which number. All three can be useful and only the third can prove it. The difference matters because the work you need depends entirely on your traffic volume: below a certain number of conversions a month, formal testing cannot reach a reliable result, and the honest answer is to apply known good practice and stop pretending the outcome was measured.
Message match, not decoration, is the first fix
The most common defect on a paid landing page is that it does not say back to the reader what they just clicked. Someone who searched for a specific service in a specific city and clicked an ad promising it should see that exact thing in the first line, above the fold, in the words they used. Instead they usually see a company positioning statement. A competent landing page specialist starts by lining up the query, the ad copy and the page headline, and fixing the mismatch before anything else. The second fix is almost always the form: too many fields, fields that ask for information the reader does not yet want to give, no explanation of what happens after they submit. The third is proof placed where the doubt occurs rather than in a testimonial band at the bottom. None of this is glamorous and all of it typically outperforms a redesign. When you are buying page-level work as part of a broader on-page programme, insist the same person owns the query, the copy and the form, because splitting them across teams is how message match gets lost.
Speed and accessibility are conversion work, not technical trivia
A page that is slow to become usable loses people before the argument starts. Google's web.dev documentation on Core Web Vitals defines the specific metrics that measure loading, interaction responsiveness and visual stability, and those metrics exist because they describe experiences that make readers leave. A landing page expert who cannot discuss them, or who ships a hero video that pushes the form below a wall of loading, is optimising something other than conversion. Accessibility works the same way. Form labels that are actually associated with their inputs, sufficient colour contrast, a sensible focus order and error messages that say what went wrong are the same changes that reduce abandonment for everyone. The ADA.gov guidance on web accessibility explains why public-facing pages are expected to be usable by people with disabilities, and in practice the fixes it points at are the ones a conversion specialist would recommend anyway. Ask a candidate how they check both before and after a change.
Testing discipline is the thing you are really buying
Ask any candidate two questions. First: how many conversions a month does a page need before you would run a split test on it? An expert has a number and will tell you that below it you should apply best practice and measure over longer windows instead of declaring winners. Second: show me a test that lost. Anyone who has run a real programme has a folder full of confident hypotheses that did nothing or made things worse, and the willingness to show you those is the strongest signal you will get in a sales call. A specialist who presents only uplifts, especially large round-number uplifts with no timeframe or sample described, is presenting marketing rather than results. Also ask what they do with a test that shows no difference, because the correct answer, keep the simpler version and move on, is one you will hear only from people who have done this at volume.
How these people are actually engaged
Landing page work is bought in three shapes. A fixed-price build, where you get a defined number of pages and a handover, suits a company with one campaign and no internal capability. A monthly retainer with a testing programme suits a business with enough traffic to learn something each month. An advisory arrangement, a few hours a month reviewing pages your own team builds, suits a company that has designers but no conversion judgement and is often the best value of the three. Beware the fourth shape, a page build bundled invisibly into an ad management fee, because it hides how much you are paying for either and gives the agency no incentive to rebuild a page that is not working. Whichever you choose, ownership matters: the pages, the analytics property and the test data should be yours, on your domain where possible, so that changing provider does not mean starting again.
Questions people ask about landing page experts
How much traffic do we need before testing is worth it?
Enough conversions, not enough visits. A page converting a handful of times a month cannot produce a trustworthy split test result in a reasonable window. Below that level, apply known good practice, fix speed and form friction, and compare over longer periods instead of declaring a winner from noise.
Should landing pages sit on our own domain or the agency's tool?
On your domain wherever possible. It keeps analytics coherent, avoids a second site competing with yours, and means you keep the pages when the relationship ends. Third-party builders are convenient for short campaigns, but ask what happens to the pages and the data if you stop paying.
Is a redesign usually the answer?
Rarely first. Message match, form length, proof placement and load performance account for most of the difference on a typical page, and each can be changed in days. A full redesign changes everything at once, which makes it impossible to know what worked.
What should a candidate show us before we hire them?
A page they wrote, the query and ad it was matched to, what they changed and what happened, including at least one test that failed. Ask how they measured it and over what period. Confident uplift figures with no method behind them are the most common thing in this field and the least informative.