You pay for every click — in euros or in SEO effort. The landing page decides what's left of it. It is the most profitable page to optimise on a site, and where mistakes are paid in cash. Here is what makes one convert, evidence included.
Quick Answer: what makes a landing page convert?
Five elements, in this order of importance:
- Continuity with the source — the message that earned the click (ad, email, post) must appear word for word above the fold.
- A one-sentence value proposition — what I get, why you rather than someone else.
- A single call to action, repeated: visible on arrival AND beyond the 50% scroll line.
- Proof — numbers, testimonials, logos, guarantees, placed at the moment of doubt, not in the footer.
- Zero parasitic friction — no scattering navigation, no interrogation-style form, no hidden cost.
A good conversion rate: 20–40% for a free sign-up, 2–6% for a B2B demo.
The structure that works, section by section
The hero: 5 seconds to earn the right to be read
Above the fold, three elements and nothing else: the headline (the promise, phrased in the visitor's words — not your slogan), a subheadline stating for whom and how, and the primary CTA. Add a visual that shows the product or the outcome — not a decorative stock illustration.
The coherence test: open the ad (or the Google snippet) and the page side by side. If the keywords of one don't appear in the other, you are paying for clicks to create confusion.
The body: handle objections in the order they arise
Past the hero, a scrolling visitor asks questions in a predictable order: does it work for my case? — how does it work? — who else uses it? — how much? — what are the risks? Each section should answer one of them, in that order, with the CTA repeated every two sections.
Social proof belongs at the moment of doubt: a customer quote near the price, logos near the promise, numbers near the how-it-works. An isolated "they trust us" block in the footer reassures nobody at the decisive moment.
The form: every field has a price
The rule: ask the minimum required for this step, collect the rest after conversion. Going from 7 to 3 fields almost always measures in double-digit percentage gains. Optional field? Remove it. "Useful for sales" information? It costs leads.
Measuring a landing page: the four numbers that matter
- Conversion rate by source. The same page converts at 30% from a customer email and 3% from cold display ads. Judge the page per source, never globally — the conversion rate guide explains why.
- The 50% scroll line. What share of traffic sees your decisive arguments? The scroll map answers; whatever lives below the line exists only for a minority.
- Dead clicks. Non-clickable elements attracting clicks: disappointed intents to repair.
- Abandoners' behaviour. Five replays of non-converting sessions say more than any report: where it hesitates, where it leaves.
The mistakes that kill a landing page
- The homepage in disguise: full navigation, three messages, five CTAs. A campaign landing should offer one exit: the conversion.
- The broken promise: the ad says "−20%", the page never mentions it. Instant bounce, budget burned.
- The self-centred text wall: "we have been leaders since 2003"… The visitor is looking for THEIR problem, not your history.
- The soft CTA: "Learn more" says neither what you get nor what it costs. "Start the free trial — no card" converts because it answers both.
- The premature form: asking for a phone number for a whitepaper is trading a PDF for a lead who will never answer.
Optimising: from intuition to test
A landing page is not optimised in a meeting. The full CRO cycle applies: measurement locates (scroll, clicks, conversions per source), replay explains, the A/B test decides. By decreasing yield, test the headline promise first, then the structure (section order, form position), then cosmetics — never the other way round.
Target your variants too: new mobile visitors from an ad don't share the frictions of returning desktop users. Behavioural segmentation turns an on-average-decent page into an excellent page per segment.
In practice with Mirage — For every landing page: conversions per source and UTM campaign, scroll and click maps, replay of non-converting sessions, and persona-targeted A/B testing — all cookieless, hence measured on 100% of visitors, including the 30–50% who refuse banners elsewhere. Free 30-day trial.
FAQ
What exactly is a landing page?
A page designed for ONE traffic source and ONE conversion goal: sign-up, demo request, purchase, download. Unlike a homepage that must serve every audience, a landing page tells one story and offers one action — which is precisely what makes it convert.
What is a good landing page conversion rate?
By goal: 20–40% for a free sign-up from qualified traffic, 2–6% for a B2B demo request, 1–3% for a direct purchase. The dominant variable is not the page itself but the match between the source's promise (ad, email) and the page content.
Should a landing page be long or short?
It depends on the commitment requested. Free, reversible action (newsletter, trial): short page, quick decision. Costly or engaging action (purchase, sales meeting): long page that handles objections. The scroll map settles it: if nobody reaches your decisive arguments, the page is too long or badly ordered.
Should the navigation menu stay on a landing page?
For a campaign landing: no — every outbound link is a leak, and conversion almost always rises when navigation is removed. For an SEO destination page, keep it: organic visitors expect to explore, and Google values internal linking.