What makes a landingStart a projectpage convert

A landing page has one job, and the discipline is in refusing everything that is not that job. Pages rarely fail because the design is ugly. They fail for three reasons that are all measurable, and all fixable.

It loads too slowly on the device that matters

Most traffic arrives on a mid-range phone over a mobile network, and that is the machine the page has to be fast on. A desktop lab score of 98 tells you almost nothing about it.

The practical target is Largest Contentful Paint under one second on that device. Getting there is rarely subtle:

• Ship the page as static HTML, without a framework runtime executing before anything appears.
• Inline the critical CSS and defer the rest.
• Self-host the fonts and preload the one used above the fold. A request to a font CDN is a round trip before the first word is legible.
• Serve images in a modern format, at the size they are displayed, with width and height set so nothing jumps.

Every additional second before the page is usable costs a measurable share of the people who arrived.

It asks for several things at once

A page with a headline CTA, a secondary CTA, a newsletter box, a chat bubble and a navigation bar is not asking for one thing. It is offering five exits, four of which are not what you wanted.

The rule that survives contact with reality: every section either moves the visitor toward the single action or comes off the page. Secondary links are text, not buttons. The navigation, on a true landing page, usually should not be there at all.

The form is part of this. Every field is a reason to stop. Ask only for what a person will genuinely use on the first call back — usually a name, a contact, and one sentence about the problem. Fields like "company size" are for your CRM's comfort, not the visitor's.

Nobody can say where it lost people

This is the one that keeps pages broken, because without it every change is a guess.

Analytics and conversion tracking go in before launch, not after the first disappointing week. What you need is modest:

• Scroll depth by section, so you know which argument they stopped reading at.
• Interaction with the form, separately from submission — how many started and abandoned, and on which field.
• The conversion itself, tied to whatever counts as a lead in your CRM.

With those three you can answer "where did we lose them" in an afternoon. Without them, you will be rewriting the headline for months because it is the only thing anyone has an opinion about.

What we do not believe moves the number

A few things get attributed more power than they have.

Above-the-fold length. Long pages convert fine when the argument is good. The rule was invented for a different web.

Button colour. It matters when contrast is genuinely poor. Otherwise it is the thing teams test when they have run out of ideas.

Urgency theatre. Countdown timers that reset on refresh and "3 people are viewing this" widgets. They lift numbers briefly and cost trust permanently, and they are trivially recognisable.

Where to start

If you have an existing page and one afternoon, do this in order: measure the load time on a real mid-range phone, count the actions the page offers, and check whether you can tell where visitors stop. In our experience, most pages fail at least two of the three, and fixing them moves more than any redesign.

Then redesign, if you still want to — but from data rather than from taste.