How to Build a Landing Page That Converts (When You're Not a Marketer)

You did the hard part. The product works, you posted about it, and people actually came — a few hundred visitors from a launch, a thread, an ad you paid for. Then you checked your signups and found four. Not four percent. Four people.

The instinct at that moment is to blame the traffic, or the product, or to conclude that marketing is some dark art you weren't born with. Usually it's none of those. It's that the page those visitors landed on was built by someone who knows the product intimately, for someone who has never heard of it — and that gap shows up in about eight seconds of a stranger's attention.

The reassuring part is that conversion isn't taste, and it isn't creativity. It's a fairly small set of mechanics that have been measured across hundreds of millions of visits, and most of them are things you can fix this afternoon without a designer. This guide walks through what a landing page that converts actually needs, what the real benchmarks are so you can judge yourself honestly, and — the part almost every guide skips — what to do when you don't have enough traffic to run a proper test.

Table of Contents

• What a good conversion rate actually is
• One page, one job
• Message match: the leak before the page
• What has to be above the fold
• The rest of the page, in the order that works
• Cold traffic and warm traffic need different pages
• Writing a headline that isn't about you
• One call to action, repeated
• Every form field costs you
• Social proof, and what actually counts as it
• Speed is a conversion feature
• Mobile is the default, not the edge case
• Video: the biggest lever and the easiest to get wrong
• How to improve when you can't A/B test
• The page after the page
• Ten mistakes that quietly cost you signups
• Frequently asked questions

What a good conversion rate actually is

Before changing anything, you need a fair scoreboard. Most founders either panic at a number that's completely normal or feel fine about one that's quietly terrible, because they're comparing against a figure they half-remember from a blog post.

The broadest reliable reference point: the median landing page converts at about 6.6%, from an analysis of hundreds of millions of visits across tens of thousands of pages. But the spread by industry is enormous, and comparing yourself to the global median is close to meaningless.

Two things to take from that. First, around 10% is genuinely good in most categories, and the best pages in favorable conditions — warm email traffic, tightly targeted campaigns — reach 15–20%. Second, your industry determines a lot of your ceiling, because it sets how complex the decision is and how much commitment you're asking for. A page asking for an email converts differently than one asking for a credit card, and neither is "better."

So set your benchmark honestly, then compare against your own page last month rather than against someone else's screenshot.

One page, one job

The single most common structural mistake is that founders build a homepage and call it a landing page. A homepage is a hub — it has navigation, an about link, pricing, a blog, social icons, and several possible next steps. A landing page is a funnel: it exists to get one specific action from one specific kind of visitor.

Every extra option is a chance to do something other than convert. The measured gap is real: pages built around a single call to action convert around 13.5%, while pages offering five or more links convert around 10.5% — a meaningful loss purely from giving people somewhere else to go.

Practically: strip the navigation, remove the footer link farm, cut the social icons that send people to a feed they'll never come back from, and decide the one action this page exists for. If you find yourself unable to choose between "start free trial" and "book a demo," that's a signal you need two pages, not one page with two buttons.

Message match: the leak before the page

Here's a leak that happens before anyone reads a word of your page. Someone clicks an ad, a post, or a link that promised a specific thing — and lands on a page about something more general. The mismatch registers instantly as "this isn't what I clicked," and they leave without ever evaluating your product.

Message match means the page continues the sentence the link started. Same promise, same words, same offer.

This matters most if you're paying for traffic, because you're buying every one of those bounces. But it applies to a launch post, a newsletter link, or a comment you left in a community. Whatever specific promise brought them, repeat it in the first thing they see. It's the cheapest conversion fix that exists — usually a headline edit.

What has to be above the fold

A visitor decides whether to keep reading in a few seconds, and they make that decision on whatever is visible without scrolling. That space has to answer three questions fast: what is this, is it for me, and what do I do next.

The most common failure here is a hero that's beautiful and uninformative: an abstract illustration, a vague slogan, and a "Learn more" button. A stranger leaves that page unable to say what you do. If you can only fix one thing on your page today, make it so a person who has never heard of you can describe your product after five seconds above the fold.

The rest of the page, in the order that works

Above the fold gets the attention, but the sections underneath are what convert a visitor who's interested-but-not-sure. There's a running order that works for most pages, and it follows the sequence of questions a skeptical person asks — in the order they ask them.

Two of those sections are the ones founders most often skip, and both are cheap to add.

"How it works" in three steps removes the fear of effort. A visitor who believes your product is good but suspects it'll take a weekend to set up will leave anyway. Three short steps — paste this, pick that, get this — converts that hesitation into an expectation of ease.

An objection section is the highest-value block on most pages, because it answers the specific doubt that's actually stopping people. You don't have to guess what those doubts are; they're already sitting in your support inbox and in the questions people ask when you demo. Put the four or five real ones on the page, answered plainly:

Cold traffic and warm traffic need different pages

The same page can convert brilliantly for one audience and terribly for another, and the variable is usually how much the visitor already knows before they arrive. A page written for people who already trust you will underperform badly on cold traffic — and vice versa.

The practical rule: the colder the traffic, the more of the page has to happen before you ask for anything. Cold visitors need the problem named, the mechanism shown, and proof supplied before the form. Warm visitors find all that patronizing and just want the button.

If you're sending several sources to one page, write it for the coldest of them, then consider a second, shorter version for your warm traffic once volume justifies it. And if you're paying for that cold traffic, the page is where most of the money is won or lost — worth reading alongside how to reduce social media ad costs, since a landing page mismatch quietly raises the price of every campaign you run.

Writing a headline that isn't about you

Founders write headlines about their product. Visitors care about their own outcome. That single swap fixes most weak headlines.

Three tests for a headline. Is it specific — could a competitor paste their name into it? If so, it says nothing. Is it about the outcome rather than the mechanism? People buy the result, not the technology. And is it in their words — the phrasing your customers actually use, which you can lift directly from support emails, reviews, and sales calls?

Category jargon is the quiet killer. You've said "workflow automation platform" a thousand times; your visitor has said it never.

One call to action, repeated

One action per page — but say it more than once. On anything longer than a screen or two, repeat the same CTA at natural decision points: after the hero, after the proof section, and at the end. Repeating the same action isn't clutter; offering different actions is.

Write buttons that name what happens next. "Submit" and "Learn more" are the two weakest labels in common use, because neither tells you what you get:

Reduce the perceived risk right under the button, too. "No credit card required," "Free forever plan," "Takes 30 seconds," "Unsubscribe anytime" — one short line addressing the hesitation your visitor actually has costs nothing and removes a real objection.

Every form field costs you

If your conversion action involves a form, this is likely your biggest single lever, and the data is unusually clear: each extra field past about four roughly halves your conversion rate.

The discipline: for every field, ask "do I need this to deliver the thing, or do I just want it?" Company size, job title, phone number and "how did you hear about us" are almost always wants. You can ask for them later, once someone is a user and has a reason to answer. Collect the minimum that lets you fulfill the promise — usually an email — and let onboarding do the rest.

Social proof, and what actually counts as it

Social proof works, and it's one of the cheapest additions available: pages that include it convert at roughly 12.5% versus 11.4% without. That's a real lift for something that costs you a copy-paste.

But not all proof is equal, and the weakest kinds are the ones founders reach for first.

If you're early and have no proof, don't fake it. Use what you honestly have: a real quote from a beta user with their name, a screenshot of genuine feedback, a founder note explaining what you built and why. Being visibly early and honest converts better than fabricated credibility, which experienced buyers spot immediately — and which becomes a permanent liability if noticed.

Speed is a conversion feature

Founders treat performance as an engineering nicety. It's a conversion input, and the numbers are steep: pages loading in under three seconds convert about 32% better than slower ones, and a single second of delay can cost around 7% of conversions.

The usual culprits on a founder-built page are unoptimized hero images, a heavy autoplaying video, too many fonts, and analytics or chat widgets loading before the content. The fastest wins: compress and correctly size the hero image, lazy-load anything below the fold, defer third-party scripts, and test on a mid-range phone over mobile data rather than on your laptop over office wifi.

Mobile is the default, not the edge case

Most of your traffic — especially from social, and especially from a launch — arrives on a phone. Yet most founder-built landing pages are designed on a wide monitor and checked on mobile at the end, if at all.

A short mobile pass that catches most damage:

Video: the biggest lever and the easiest to get wrong

Video on a landing page is one of the largest documented conversion levers available. In a well-known controlled case study, the same page converted at 6.5% without video, 11% with the video embedded (a 69% lift), and 13% with it in a lightbox (a 100% lift). Broader analyses put typical lifts in the tens of percent, with the upper end near 80%.

It works because it solves the exact problem a founder-built page has: it shows the product instead of describing it, in the seconds where a stranger decides whether to care.

But — and this caveat matters more than the headline number — the lift is not automatic. A long, off-topic, or slow-loading video makes the page worse, and a heavy video file can wipe out its own benefit through the speed penalty described above. What separates the pages that gain from the ones that lose:

The practical obstacle is that producing a tight, on-brand demo video is exactly the kind of task a founder postpones forever. That's the part worth removing: with ArtFlyAI you paste your site's URL and get a short, captioned, on-brand video built from your own page — colors, product, language — in about a minute, which is short enough to fit the rules above rather than fight them. If you want the craft side of what to actually show, our guide on how to make a demo video for your app covers structure and hooks in depth.

How to improve when you can't A/B test

Every conversion guide tells you to A/B test. Almost none of them mention that a valid test needs far more traffic than a young product has — you generally need hundreds of conversions per variant to detect the kind of difference you're looking for. With 300 visitors a month, an A/B test will produce a number that looks like an answer and isn't.

So here's the honest sequence for low traffic. Do these in order; don't skip to testing.

What replaces testing at low volume is qualitative evidence, and it's often better anyway:

Five strangers telling you they couldn't work out what your product does is more actionable than a statistically meaningless test — and you can act on it the same day.

The page after the page

Conversion doesn't end at the button. The screen someone sees immediately after signing up is one of the most neglected pages on the internet — usually a bare "Thanks! Check your email" — and it's arriving at the single highest-intent moment you will ever get from that person. They just said yes. They're paying attention. And most pages respond by doing nothing with it.

That moment is worth designing, because it decides whether a signup becomes a user:

Two things reliably help. First, reduce the gap between yes and value — if the promise was a template, put a download button on the confirmation screen rather than sending them to hunt through email. Every extra step between intent and payoff loses people who genuinely wanted it.

Second, use the moment while attention is highest. This is the best time to ask for the thing you'd normally never get: a follow, a calendar booking, a two-question survey about what they were hoping for. A survey answer here is also the cheapest source of headline copy you'll ever find, because it's people describing your value in their own words at the exact moment they decided to try you.

Ten mistakes that quietly cost you signups

Ordered by how much they typically cost.

1. Using your homepage as a landing page. Many exits, no single job. The most expensive structural error on this list.

2. A hero that doesn't say what you do. Beautiful, abstract, and useless to a stranger.

3. Message mismatch. The page doesn't continue the promise that earned the click, so people leave before evaluating anything.

4. Asking for too much information. Every field past four roughly halves conversion. Ask for what you need to deliver, not what you'd like to know.

5. Slow load. Past four seconds, conversion collapses. Usually an oversized hero image.

6. Competing calls to action. Two equal buttons is a decision, and decisions cause exits. One action, repeated.

7. Writing about features instead of outcomes. You're proud of how it works; they care what changes for them.

8. No proof at all. A page with zero evidence asks for trust it hasn't earned. Even one honest named quote helps.

9. Ignoring mobile. Designed on a monitor, experienced on a phone — where most of your traffic actually arrives.

10. Rewriting the whole page every week. Constant churn means you never learn which change did what. Change one meaningful thing, give it enough traffic, then move on.

Frequently asked questions

What is a good landing page conversion rate?

Around 10% is genuinely good in most industries, and the best pages reach 15–20% with warm, well-targeted traffic. The all-industry median is roughly 6.6%, but the spread is wide — SaaS averages near 3.8%, e-commerce around 2.5–3.5%, and financial services or lead-gen pages often exceed 8%. Compare against your industry and against your own previous numbers rather than a global average.

How many form fields should a landing page have?

Three to five is the practical sweet spot, and fewer is usually better. A single email field converts near 12.4% while six or more fields drops to around 3.1% — each field past about four roughly halves your rate. Cutting a form from eleven fields to four has been measured to increase conversions by roughly 120%. Ask only for what you need to deliver the thing you promised.

Should a landing page have navigation?

Generally no. A landing page has one job, and navigation gives visitors other places to go. Pages with a single call to action convert around 13.5% versus about 10.5% for pages with five or more links. Keep the nav on your homepage; strip it from pages built to convert.

Does video actually increase landing page conversions?

Yes, substantially, when done right. In a well-known case study the same page went from 6.5% without video to 11% with it embedded and 13% with it in a lightbox — lifts of 69% and 100%. But it isn't automatic: a long, off-topic, or slow-loading video can lower conversion. Keep it to 30–90 seconds, show the product doing the thing the page promises, add captions, and compress it.

How fast does my landing page need to load?

Under three seconds, and under two is better. Pages loading in under three seconds convert about 32% better than slower ones, and each additional second costs roughly 7% of conversions. Past four seconds conversion drops sharply. The usual fix is compressing the hero image and deferring third-party scripts.

How much traffic do I need to A/B test?

More than most young products have — realistically hundreds of conversions per variant to detect a meaningful difference. Below roughly 10,000 visitors a month, A/B tests produce numbers that look authoritative and aren't. Fix the known fundamentals first, then use session recordings and user conversations, and save testing for when you have the volume.

What should go above the fold?

Five things, in this order: a headline stating the outcome in the visitor's words, a one-sentence subhead on how it works or who it's for, visual proof of the product doing the thing, a single call to action naming the outcome, and one trust signal. The test is whether a stranger can describe your product after five seconds.

How do I write a landing page headline?

Write about the visitor's outcome, not your product's mechanism, using words your customers actually say. Three checks: could a competitor paste their name into it (too generic), is it about the result rather than the technology, and is it free of category jargon you use but your visitor doesn't. Support emails and reviews are the best source of the exact phrasing.

What's the difference between a landing page and a homepage?

A homepage is a hub designed for exploration — navigation, pricing, blog, about, several possible next steps. A landing page is a funnel built for one action from one kind of visitor, with everything else removed. Using your homepage as a landing page is the most common and most costly structural mistake.

What if I have no testimonials or customers yet?

Use what you honestly have and don't fabricate anything. A real named quote from one beta user, a screenshot of genuine feedback, or a short founder note explaining what you built and why all convert better than invented testimonials — which experienced buyers recognize immediately and which permanently damage trust if noticed.

Where should I put the call to action?

Above the fold, and then repeated at natural decision points — after your proof section and at the end of the page. Repeating the same action is good; offering different actions is what hurts. Name the outcome on the button ("Create my first video" beats "Submit") and add one short risk-reducing line underneath.

How often should I change my landing page?

Change one meaningful thing at a time and give it enough traffic to mean something before changing the next. Rewriting the whole page weekly guarantees you'll never know which change mattered. At low traffic, a monthly cadence of one substantial improvement plus qualitative feedback beats constant churn.

Summary

A landing page that converts isn't a design achievement — it's a sequence of mechanical decisions, most of which you can fix today. Give the page one job and remove every other exit. Continue the promise that earned the click. Put the outcome, the proof, and one clear action above the fold, written in your visitor's words rather than your product's vocabulary. Ask for as little information as you can get away with, because every field past four roughly halves your conversion. Show honest proof. Keep it under three seconds and check it on a real phone. Add a short, captioned video showing the product doing the thing — the single biggest documented lever, as long as it stays short and fast.

And if you don't have the traffic to test properly, don't pretend you do. Fix the known fundamentals, then go watch five strangers try to understand your page. That hour will teach you more than any test you're not big enough to run yet.

Want the video part handled? Paste your landing page URL into ArtFlyAI and get a short, captioned demo built from your own page in about a minute.