Every founder I talk to thinks their landing page problem is a design problem. It rarely is. I've rebuilt dozens of landing pages for clients, and in almost every case the original wasn't ugly; it was unclear. Good landing page design isn't about picking a nicer font or a trendier colour palette. It's about removing every reason a visitor has to hesitate, scroll away, or open a new tab and forget you exist. Get that right and the aesthetics take care of themselves.
The one goal a landing page actually has
A website has to do a dozen jobs. A landing page has to do one.
That sounds obvious written down, but it's the single most common mistake I see. A landing page built for a paid ad campaign will have a pricing table, three testimonials, a blog teaser, a newsletter signup, and a footer with eleven links to other pages. Every one of those elements is a door out of the conversion you're trying to get.
Before I touch a single visual element on a landing page brief, I ask the client one question: what's the one action you want someone to take? Book a call. Start a trial. Add to basket. Download the report. Everything on the page should either build the case for that one action or get out of the way.
If you can't answer that question in a single sentence, you don't have a landing page brief yet — you have a homepage brief wearing a landing page's clothes.
Why most landing pages fail before design even starts
This is the bit nobody wants to hear: most landing pages fail in the brief, not in Figma.
I'll get sent a page that's converting badly and the client wants a "redesign." Nine times out of ten, when I read the copy, the problem isn't visual at all. The headline describes the product instead of the outcome. The value proposition is buried in paragraph three. There's no clear answer to "why should I care about this in the next five seconds."
Design can't fix a message that isn't there. It can dress up a weak argument, but weak arguments don't convert no matter how well they're typeset. So before any layout decisions get made, I want three things nailed down: who's landing on this page, what they already believe, and what single sentence changes their mind. Everything downstream, headline, subhead, structure, CTA copy, flows from that.
The anatomy of a page that actually converts
Once the message is sorted, the structure is fairly predictable. It's not exciting, and that's rather the point. Conversion isn't the place to get experimental.
Headline. States the outcome, not the feature. "Cut your invoice processing time in half" beats "AI-powered invoicing software" every time, because the first sentence answers "what's in it for me" and the second makes the visitor do the maths themselves. Most people won't bother.
Subheadline. Does the job the headline can't. It adds the specific detail, the "how," or the qualifier that makes the promise credible. If your headline is bold, your subhead should be believable.
Hero visual. Should show the thing, not represent the thing. Abstract illustrations of gears and lightbulbs tell a visitor nothing. A screenshot of the actual product, a photo of the actual result, or a short demo clip does far more work in the same amount of space.
Primary CTA. One action, repeated, not competing with a second option of equal visual weight. If you offer "Start free trial" and "Watch demo" as equally sized buttons, you've just made your visitor choose between two decisions instead of making one. Pick a primary and a secondary, and design them so the hierarchy is obvious at a glance.
Social proof. Placed close to the point of decision, not exiled to a testimonials section at the bottom nobody scrolls to. A single specific result, a name, a number, a real outcome, does more than five generic five-star quotes.
Form or checkout friction. Every field you ask for is a small tax on conversion. If you don't need a phone number to book a discovery call, don't ask for one. I've watched conversion rates jump meaningfully just from cutting a form from seven fields to three.
Copy comes before colour
I say this to clients more than almost anything else: write the words before you design the page. Not placeholder copy, not "lorem ipsum, we'll swap it later." Actual, final copy.
The reason is simple. Layout is a response to content, not the other way round. If you design a beautiful hero section first and then try to squeeze your value proposition into it, you'll end up cutting the sentence that actually persuades people because it doesn't fit the box. I've seen this happen constantly with in-house teams working against a template — the template dictates the message instead of the message dictating the layout.
Get the copy right first. Headline, subhead, the three or four supporting points, the CTA text, the objection you need to overcome. Then design around it. It's a less glamorous way to work, but it's the difference between a page that looks finished and a page that actually persuades.
The above-the-fold obsession is overrated
There's a piece of landing page folklore that refuses to die: "everything important must be above the fold." I understand where it comes from, but it's outdated advice built for an era before people were used to scrolling on their phones a hundred times a day.
What matters isn't cramming everything into the first screen. What matters is that the first screen earns the scroll. If your headline, subhead, and hero image make someone curious enough to keep going, they'll go. If your above-the-fold section is trying to close the sale in one glance, it usually ends up cluttered and unconvincing instead.
A landing page's first screen has one job: earn the next screen. It doesn't need to close the deal; it needs to stop the scroll.
Save the proof, the detail, and the objection-handling for the sections that follow. Let the page build a case in order, the way a good salesperson would in conversation, rather than trying to dump the entire pitch on someone in the first three seconds.
Trust signals that actually move people
Not all social proof is equal, and this is where a lot of landing pages waste good space.
Logos of companies you've worked with are fine, but they're passive. A visitor glances and moves on unless the logo is genuinely recognisable to them. What moves people is specificity: a named person, a real result, a timeframe. "Sarah cut her onboarding time from three weeks to four days" does more work than a wall of five-star ratings with no context.
If you're B2B, industry-specific proof outperforms generic proof. A prospective client in logistics cares far more about a result from another logistics company than a glowing quote from an unrelated sector, even if the unrelated quote is more enthusiastic. Relevance beats volume.
And if you genuinely don't have strong proof yet (you're early, you're new), don't fake it with vague claims. A confident, specific promise from the founder, in their own voice, often outperforms weak manufactured testimonials. People can tell the difference.
Mobile isn't an afterthought, it's the majority
I still see landing pages designed on a widescreen monitor first, with mobile treated as a compression exercise afterwards. For most B2C traffic and a good chunk of B2B traffic arriving from ads or social, mobile is the majority of your visitors, not the edge case.
That changes real decisions, not just spacing. Long headlines that look confident on desktop wrap awkwardly and lose punch on a phone. Forms that feel fine with a mouse become genuinely annoying to fill in with a thumb. CTAs that sit comfortably in a desktop hero section can end up below the fold entirely on mobile if you don't design for it deliberately.
Design mobile first, or at the very least design mobile properly, as its own layout decision, not a scaled-down afterthought. It's usually where the biggest, easiest conversion gains are sitting unclaimed.
Testing beats guessing
Everything above is a strong starting point, not gospel. The honest answer to "what converts" is: it depends on your audience, your offer, and your traffic source, and the only way to know for certain is to test.
That doesn't mean you need a six-week A/B testing programme before you launch anything. It means treating your first version as a hypothesis, not a finished product. Watch where people drop off. Change one meaningful thing at a time, the headline, the CTA copy, the form length, and see what moves. Small, deliberate changes tell you far more than a full redesign every few months based on a hunch.
The bottom line
Landing page design isn't about making something look impressive. It's about removing friction between a clear message and a single action, and doing it with enough clarity that a stranger understands your offer in seconds, not minutes. At Slate, every landing page starts with that one-sentence brief: who's landing here, and what single action are we asking them to take, before a single pixel gets placed.
Ready to simplify your design process? See plans.