Most Landing Page Problems Are Message Match Problems
Why the fix is almost never the page design
When a campaign converts badly the investigation usually starts with the page: the layout, the button colour, the length of the form. Those are measurable and they are rarely the binding problem. The binding problem is that the visitor arrived with a specific expectation and the first screen did not confirm it.
That decision takes about a second and happens before any of the design choices get a chance to matter. It is also the cheapest thing on this list to fix, because it means changing a heading rather than rebuilding a template.
The gap between what the ad promised and what the page says is where budget disappears.
The diagnosis
A visitor arrives with a specific expectation created by the ad. If the page headline does not confirm it within a second, they leave, and no amount of design work changes that.
This is why high click through rates sometimes come with terrible conversion rates.
Fixing it
Match the headline to the ad’s promise in the ad’s words. Keep the offer visible without scrolling. Remove navigation that invites wandering. Make the next step obvious and singular.
Then test, because the version you prefer is not always the version that works.
Measure it properly
Compare by campaign and device rather than looking at a site average. One bad page can hide behind a good overall number.
Measuring the gap
Put the ad and the page side by side on a phone screen and read only what is visible without scrolling. If the specific promise in the ad does not appear in those words on that screen, the visitor has to work out whether they are in the right place, and a meaningful share of them decide they are not. The decision takes about a second and it happens before anything else on the page gets a chance.
The expensive version of this is a campaign with ten ad groups pointing at one generic page. Each group promised something different and the page answers none of them specifically. The fix is rarely ten pages; it is usually one page whose first screen changes to match the promise, with the rest shared.
Closing it
- List every ad group and the exact promise its best-performing ad makes.
- Open the landing page for each and check the promise appears above the fold, in the same words.
- Where it does not, change the page heading rather than the ad, because the ad earned the click.
- Keep one call to action on that first screen, matching the stage the visitor is at.
- Re-check on a phone, at the width most of your traffic actually uses.
Matching the keyword instead of the promise
Dynamically inserting the search term into the heading produces grammatically strange pages that match the string and miss the point. What the visitor is checking is whether the thing they were promised is here, which is a claim rather than a phrase.
Common questions
Do we need a page per ad group?
No. One page with a swappable first screen usually performs as well and is far cheaper to maintain.
How much copy should be above the fold?
The promise, one supporting line, and one action. Everything else can wait.
Does this apply to organic landing pages?
The same logic applies to the query rather than the ad: the page should confirm, in its first screen, that it answers what was searched.
What if the ad overpromises?
Then fix the ad. A page that honestly cannot deliver the promise converts badly and generates refunds.
