Landing page design

One page, one action

The most common landing page mistake is trying to serve several audiences at once. A page with three calls to action has no call to action, because the visitor has to make a decision before they can act and most will decline.

Decide the single action before anything is designed. Book a consultation. Request a quote. Start a trial. Download the specification. Everything on the page then either supports that action or is removed, which is a surprisingly effective editing rule.

Message match

A visitor arrives holding an expectation created by the ad they clicked. If the page does not confirm that expectation within a second, they leave, and they leave before reading anything.

So the headline should echo the promise of the ad, in similar words. If you advertise a specific treatment, the page is about that treatment, not about your practice generally. If the ad names a price, the page confirms it immediately. This sounds obvious and it is the most common reason paid traffic underperforms.

Speed, on the connection they actually have

Paid traffic is disproportionately mobile and frequently on an unreliable connection. Every additional second costs conversions, and the loss is heaviest at the start.

Landing pages should be the fastest thing you own: minimal CSS, almost no JavaScript, images sized and lazy-loaded below the fold, fonts preloaded, and a strict limit on tracking scripts. Sub-second is achievable for a page like this and it is the cheapest conversion improvement available.

The structure that works

Order matters more than most page elements. Above the fold: what this is, who it is for, and the action. Then the reason to believe it, which means specifics, numbers and named proof rather than adjectives. Then the objections, answered directly, because unanswered objections stop people quietly. Then how it works, so the next step feels safe. Then the action again, because people scroll to the bottom before deciding.

Forms are where conversions leak

Every field costs completions. For a first contact you need a name, a way to reach them and enough context to reply usefully. Everything else can be gathered on the call.

Beyond field count: label fields visibly rather than relying on placeholders, use the correct input types so mobile keyboards behave, show errors next to the field rather than at the top, and never clear a form on a validation failure. These are small and they compound.

Trust signals that mean something

Generic reassurance is ignored. Specific reassurance is read. A named person with a photograph, a response time you actually meet, a genuine number, and a plain statement of what happens after they submit. “I read every enquiry myself and reply within 24 hours” outperforms a row of badges.

Measurement before optimisation

Conversion tracking is verified before launch, not discovered to be broken two months later. That means testing the conversion event end to end, confirming the source is captured at submission, and checking the numbers reconcile with what actually arrives in the inbox or CRM.

After that, testing is only worth doing with enough traffic to reach a real conclusion. Below a few hundred conversions a month, most tests are noise, and you are better served by fixing the obvious problems than by running a test you cannot read. I will tell you which situation you are in.

What happens after they submit

The page after the form is the most neglected asset in paid marketing. A generic thank-you message wastes the one moment when a visitor is most engaged with you, and it leaves them uncertain about what they have just started.

Tell them what happens next and when: who will contact them, on what channel, within what time. Give them something to do in the meantime, whether that is a genuinely useful document, a case study relevant to what they asked about, or the option to book a slot directly instead of waiting for a reply. And make sure the confirmation email arrives immediately and reads like it came from a person, because a silent gap after submission is where doubt sets in.

This is also where the conversion event should fire, on the confirmation view rather than on the button click, so your measurement counts completed submissions rather than attempts.