Landing page

A visitor arriving from an ad is carrying an expectation the ad created, and they check it against the page within a second or two. If the headline confirms what they were promised, they stay. If it does not, they leave, and the click is spent whether or not anybody reads the rest.

A homepage almost never confirms it, and the reason is structural rather than a design failing. A homepage is built to serve every audience the business has: existing customers, job applicants, journalists, people looking for the phone number. Serving all of them means leading with something general, and something general is not the answer to the specific question the ad raised.

The practical version of message match is narrow: the phrase in the ad should be recognisably present in the first heading on the page. Not identical, but close enough that somebody scanning sees they arrived in the right place.

This is also why one page per campaign theme beats one page for the whole account. A page that answers three questions answers none of them first.

The difference between a landing page and a website page is that a landing page has a single job. Everything on it either moves somebody toward one action or is a distraction from it.

Navigation is the usual first casualty. A full site menu at the top of a landing page offers a dozen alternatives to converting, all of which look more casual than filling in a form. Most landing pages perform better with it removed or reduced to a logo.

Form length is the second lever and the one with a real trade. Every field costs completions, and every field removed costs qualification. The right length depends on what the sales team needs to act, which is a question for the client rather than a rule of thumb.

The exception worth remembering is that a considered purchase sometimes needs the extra information a fuller page provides. Stripping a page to a headline and a button suits an impulse offer and starves a decision that requires evidence.

Cold paid traffic is the least patient audience a site receives. They have no relationship with the brand, no particular reason to wait, and the alternative is one tap away in a results page they were just looking at. A slow page loses them before it gets to make its argument.

Core Web Vitals give a measurable target rather than an opinion: Largest Contentful Paint under 2.5 seconds, Interaction to Next Paint under 200 milliseconds, Cumulative Layout Shift under 0.1, assessed on real visits at the 75th percentile.

On paid search there is a second payment. Landing page experience is one of the three Quality Score components, so a slow or poorly matched page raises Ad Rank cost across every keyword pointing at it. The page is not only losing conversions; it is inflating the price of the clicks that reach it.

That double effect is the strongest available argument when asking a client to prioritise page work, because it converts a design request into a media cost.

In most agency relationships the landing page sits on the wrong side of the line. The agency runs the advertising and can change it in an hour. The page belongs to the client, lives in their CMS, and changes when their developer has capacity, which may be next quarter.

That produces a recognisable failure. The agency identifies the page as the constraint, mentions it in a monthly report, nothing specific was requested so nothing happens, and months later the agency is accountable for performance limited by something it was never able to touch.

The fix is to treat it as a dependency in writing from the start: which pages the campaigns depend on, who can change them, and what the turnaround is. That converts a recurring complaint into a decision the client has made.

Where the client cannot move, the alternative worth proposing is a landing page tool the agency controls, on a subdomain. It costs a little and it removes the bottleneck entirely, which is usually cheaper than a quarter of constrained performance.

Landing pages are where agencies most often break their own ability to learn. A new page ships the same week the ad copy changes and the budget moves, results improve, and nobody can say which of the three did it or whether it will hold.

The discipline is ordinary experimental practice: change one thing, give it enough traffic to resolve a difference worth acting on, and estimate that traffic requirement before launching rather than discovering afterwards that the test could never have concluded.

On most client accounts the honest finding is that the traffic cannot support testing small changes. That is worth knowing, because it redirects effort toward changes large enough to detect, which are usually the offer or the page structure rather than a button colour.