Should a Landing Page Be Long or Short?

Should a landing page be long or short?

Long enough to answer every question that would stop someone acting, and not a sentence longer. That sounds like a dodge, so here is the specific version: for most B2B pages the evidence points at the middle, somewhere in the low hundreds of words, written simply. Very short and very long both underperform.

This argument usually gets decided by whoever has the strongest opinion in the room. The designer wants it short and clean. The founder wants every objection handled. Neither is arguing from evidence, and there is actually quite a lot of evidence available.

So here is what the research says, where it stops being useful, and how we would decide for a real page.

How far down the page do people actually look?

Further than the old advice assumes, and not as far as long page advocates hope. Nielsen Norman Group published an eyetracking study on 15 April 2018, analysing over 130,000 eye fixations from 120 participants across thousands of sites. It found users spent about 57 percent of their page viewing time above the fold.

The more useful number is the next one. NN/g found 74 percent of viewing time was spent in the first two screenfuls, up to 2160 pixels. So attention does travel below the fold, and then it falls off a cliff.

Behaviour has shifted over time. The same research noted that an equivalent 2010 study found 80 percent of viewing time above the fold, so people have become more willing to scroll. But NN/g was clear that the sharp decrease in attention after the fold remained the same pattern in 2018 as in 2010. People scroll more. They do not read more. Our piece on above the fold design covers what to put in that first screen.

Does more copy actually convert better?

Up to a point, then it stops. Unbounce's 2024 Conversion Benchmark Report, drawn from over 57 million landing page conversions, found that landing pages with between 250 and 725 words of copy convert best, at a median conversion rate of 3.8 percent in SaaS.

That range is worth sitting with, because it is wider than a minimalist would like and far shorter than a long form sales page. It is roughly one to three screens of real content, which lines up neatly with what the eyetracking data says people will actually look at.

For context on the baseline, Unbounce put the SaaS median conversion rate at 3.8 percent against an all industry median of 6.6 percent. SaaS is a harder sell than most categories, which is exactly why the temptation to keep adding copy is strongest there.

Why does reading level matter more than length?

Because it changes the outcome far more dramatically. The same Unbounce report found that simple copy converts 514 percent better than overly difficult copy in SaaS. Pages written at a fifth to seventh grade reading level achieved a 12.9 percent conversion rate, against 2.1 percent for professional level copy.

Read those two numbers again. The gap between simple and complex writing is an order of magnitude larger than anything the length debate produces. You can win far more by rewriting a 400 word page in plain language than by agonising over whether it should be 300 words or 600.

Unbounce also found that high converting pages use between 50 and 140 difficult words, defined as words of three syllables or more. That is a usable editing target. Count the long words, not the total words.

When is a short page the right answer?

When the offer is familiar, cheap, or reversible, and when the traffic is already warm. Somebody arriving from your own email, who already knows the product, clicking through to book a demo does not need to be sold again. They need the form.

Short also wins when the ask is small. A newsletter signup, a free tool, a template download. The mental cost of the decision is low, so the page only needs to answer what is this and what happens next.

The mistake with short pages is not brevity, it is omission. A short page that leaves out pricing, or who it is for, or what happens after you submit, is not concise. It is incomplete, and the reader fills the gap with suspicion.

When does a long page earn its length?

When the decision is expensive, unfamiliar, or has to be justified to someone else. A 40,000 dollar annual contract that a buyer must defend to their finance team needs material that makes that defence possible. That material has to live somewhere.

Long also earns its place when the audience is mixed. A page serving both a technical evaluator and an executive sponsor genuinely has two jobs, and trying to do both in 200 words produces something that satisfies neither. Our notes on designing for two audiences cover that specific problem.

The honest test is whether each section answers a question a real buyer asked. If you can name the person who asked it, keep it. If it exists because the page looked short, cut it.

What does a long page actually cost you?

Weight, mostly, and weight costs you the people who never see any of it. Long pages accumulate images, embedded video, and animation, and those are what break performance rather than the words.

The field data is not encouraging. The HTTP Archive Web Almanac's 2025 performance chapter, using Chrome UX Report data with a primary analysis from July 2025, found good Largest Contentful Paint on just 62 percent of origins on mobile, and only 48 percent of origins passing all three Core Web Vitals there.

A long page also costs maintenance. Every claim is something that goes out of date, every section is something a future editor has to understand, and pages that nobody maintains quietly become wrong. Short pages age better because there is less of them to rot.

How do you structure a long page so it still works?

Treat it as a stack of short pages rather than one continuous argument. Each section should stand alone, open with a clear heading, and answer its own question in the first line or two. Somebody scanning should be able to stop anywhere and understand where they are.

Repeat the call to action rather than saving it for the end. If attention drops sharply after the second screenful, as the NN/g research indicates, then a single button at the bottom is reachable only by the small group who got there.

Front load the substance. The strongest proof, the clearest statement of who it is for, and the price if you publish it should all be high on the page. Saving your best material for the end is a print convention, and the data says the web does not work that way. Our guide to landing page design covers the section order we start from.

How do you find out which one your audience wants?

Test it properly or do not claim to know. This is one of the few design questions where a clean test is genuinely achievable, because the pages are separate URLs and the outcome is a countable event.

Run it long enough to mean something. A week of traffic on a B2B page is almost never enough to distinguish a real difference from noise, and the temptation to call it early is how teams end up confidently wrong. If your volume cannot support a test, accept that and decide on judgement instead of pretending.

Where testing is not viable, use behaviour instead. Scroll depth tells you whether anyone reaches the lower sections. If 80 percent of visitors never see the bottom third of the page, the argument about whether that content converts is already settled.

What would we do?

Start in the 250 to 725 word range that Unbounce's data points at, write it at a reading level a tired person can absorb, and put the strongest proof in the first two screenfuls. Then add length only where a named objection demands it.

If you inherit a very long page that performs, do not shorten it on principle. Measure scroll depth first. Cut what nobody reaches, simplify the language throughout, and you will usually get the benefit of a short page without losing the substance of a long one.

If you are arguing about page length internally and want a second opinion grounded in what your own visitors do, we are happy to take a look. You can find us at phoenix.studio.