Long-form vs short-form landing pages: how long should a presell page be?

Page length should follow the questions a reader still needs answered. Start with the shortest page that covers the offer and its limits. Compare it with a fuller version only under a planned test. Keep the destination and tab behavior fixed so the comparison stays interpretable.

Illustration of two editors comparing short and long versions of the same sales argument.
On this page 10 sections
  1. The gap decides the length
  2. Five factors that make a page longer or shorter
  3. What our nine example pages show
  4. The ad is part of the length
  5. Long pages are built for skimmers
  6. When short wins
  7. How to test length
  8. Are you testing length or changing the navigation task?
  9. How long should your page be?
  10. Equal scroll percentages can refer to different explanations

A long page and a short page are two treatments to evaluate, not two proven recipes. Identify the reader’s unresolved questions and answer them without padding. Someone comparing an unfamiliar offer may need more explanation than a returning buyer, but that distinction does not supply a universal word count or a known purchase result.

Use this guide to plan what the page needs to explain, then test the actual route. More words can add useful evidence or merely add work; fewer words can improve focus or remove information a reader still needs.

85median words in a presell adagainst 70 for product-page ads1
18.4%of presell ads run past 250 wordsagainst 8.6% for product pages1
28%of the words on a page, at mostread during an average visit2

The gap decides the length

Before buying, a reader has to accept a chain of beliefs. Every belief they do not hold yet is a section of your page:

  1. I have this problem, and it is worth solving.
  2. It has a cause I had not understood.
  3. This kind of solution acts on that cause.
  4. This product is the right one of that kind.
  5. It works for people like me.
  6. It is worth the price, and buying it is safe.

An offer page for warm visitors only has to cover the last one or two. An advertorial for a cold audience covers all six. That is why the right length cannot be decided in the abstract: it is the sum of what your particular reader is missing.

| The reader arrives believing... | The page must still cover... | Illustrative planning length, not a benchmark |

|---|---|---|

| Nothing yet, the problem feels normal | All six beliefs | Long: 1,500 to 3,000 words |

| "I have a problem, it is my mattress" | Cause, solution, product, proof, price | Long: 1,500 to 2,500 words |

| "I need one of these, but which?" | Product, proof, price | Medium: 800 to 1,500 words |

| "I like this product, is now the time?" | Price, offer, safety | Short: one to three screens |

These ranges are editorial starting points, not measured conversion thresholds or a requirement to reach a word quota.

This is the customer awareness idea, first laid out by Eugene Schwartz in Breakthrough Advertising5, applied to word count: the less aware the reader, the more ground the page has to cover.

The presell format picker turns these factors into a ranked shortlist of formats.

Five factors that make a page longer or shorter

1. Awareness. The biggest factor, covered above. Cold traffic from a feed is usually unaware or problem-aware; retargeting and email traffic is often product-aware.

2. Price and risk. A $19 impulse buy can be justified in a few lines. A $739 device, a subscription or anything that touches health or the body raises more questions, and each question needs an answer on the page.

3. A new mechanism. If your product works in a way people have never heard of (a knee spacer instead of a new mattress, silica crystals that keep urine from staying liquid), the explanation takes room. If everyone understands it at a glance, skip the lecture.

4. The number of serious doubts. List what buyers ask before they pay: does it work, will it work for me, is it safe, why this price, what if it does not fit. Each real doubt is a paragraph. Imaginary doubts are padding.

5. How much the ad already said. The ad is part of the page. A 300-word ad that tells the first half of the story lets the page start in the middle.

What our nine example pages show

Here is how long the nine pages in our examples gallery are, counting all visible text, questions sections included:

Interactive

Words on each example page

Visible text, rounded to the nearest 50 words. The quiz and the calculator also show result text that depends on the answers, not counted here.

  • Advertorial (Cellsius)~2,500
  • Listicle (Pacha)~1,950
  • Diary (TrèsCalme)~1,600
  • Comparison (Megelin)~1,550
  • Discovery article (MyTamra)~1,500
  • Founder letter (Pratikae)~1,250
  • Calculator (Wild)~1,050
  • Offer page (Thrivin)~850
  • Quiz landing (True Nutra)~550

Counted on the published pages. Six are in French, which runs slightly longer than English for the same content.

The order follows the gap. The two longest pages, the advertorial and the listicle, speak to readers who blame the wrong cause: they must dismantle a belief before they can sell anything. The offer page speaks to visitors who already like the product, so it opens with five words of benefit and the deal. The quiz page is short on purpose, because the visitor's own answers will do the convincing.

Notice also what the long pages do not contain: company history, a list of every feature, three paragraphs of introduction. They are long because the argument is long, not because the writer was.

In our research database of 1.29 million ads collected from the public Meta Ad Library, ads that send traffic to a presell page are longer than ads that send it to a store. Look at the share of real long-form ads, over 250 words:

Interactive

How long the ads are, by destination

Ads that started running between April 11 and October 11, 2026.

Over 250 words

Share of ads with more than 250 words of copy.

  • Listicle27.7%
  • Advertorial19.8%
  • Presell pages, all18.4%
  • Quiz funnel11.0%
  • All ads10.6%
  • Product page8.6%
  • Collection page4.4%

Median words

Half of the ads are shorter, half are longer.

  • Listicle113 words
  • Advertorial93 words
  • Presell pages, all85 words
  • Product page70 words
  • All ads65 words
  • Quiz funnel49 words
  • Collection page39 words

Words of primary text in the ad. App-store ads are excluded.

Presell advertisers treat the ad as chapter one1. Listicle ads are the longest of all: more than one in four runs past 250 words. Quiz ads go the other way (median 49 words): their only job is to make the question irresistible, and the quiz does the talking.

The practical rule: plan the ad and the page as one piece of writing. If the ad tells the story's opening, the page can start at the turning point; if the ad is a single line, the page must do all the work. Either way, keep the message match: the page continues the ad, it does not restart it.

Long pages are built for skimmers

Long-form does not mean people read every word. Jakob Nielsen's analysis of browsing data found that on an average visit, people have time to read at most 28% of the words on a page, and 20% is more likely2. Eye-tracking studies by the Nielsen Norman Group show people scanning in patterns, the best known being the F-shape: they read the first lines, then mostly the starts of lines and subheads further down3. And the first seconds decide whether they stay at all: a page has to show its value quickly to earn the scroll4.

So a long page is really two pages on top of each other: the full argument for the people who read, and a fast version for the people who scan.

Long and unreadable

  • Subheads like "Our story" and "Why us"
  • Paragraphs of eight lines on a phone
  • The product's benefits listed as adjectives
  • One button, at the very bottom
  • Images chosen to decorate

Long and skimmable

  • Subheads that are full claims: read alone, they tell the story
  • One idea per paragraph, two to four lines on a phone
  • Numbers, dates and names instead of adjectives
  • A button after the reveal, after the proof and at the end
  • Each image proves something the text says

A good test: read only the headline, the subheads, the bold lines and the captions. If that skeleton makes the full case, the page is long and fast. If it does not, the page is not too long, it is badly built.

When short wins

A shorter treatment can be worth testing when the reader already understands the offer. Retargeting, email or a prior video view can suggest that context, but none guarantees that the reader has resolved every doubt. Preserve the information needed for the actual decision.

The Thrivin offer page is a good model: "One piece. Held in. Squat-proof." and the offer in a box, on the first screen. But short does not mean empty. In about 850 words it still explains why now, what makes the fabric different, how the offer works item by item, what four-star reviewers say, and what happens if it does not fit. A short page skips the parts the reader already believes; it does not skip the doubts they still have.

How to test length

Compare length only under a planned experiment with enough information for the question. Keep the ad, offer, destination and opening behavior consistent, and define the outcome before interpreting a result:

  • Product-page click-through is a route diagnostic, not proof of persuasion or profitable orders. Keep its numerator and denominator explicit.
  • End-to-end purchases per assigned visitor address a different question from product-page click-through. Include cost and contribution when the decision concerns economics.
  • Scroll and exit observations can locate a point to investigate. They do not explain the cause by themselves; inspect the content, layout, loading and route before changing the opening.

Plan the sample size before you start, and do not stop when one version pulls ahead early. Our guide to A/B testing presell pages has calculators for both.

Are you testing length or changing the navigation task?

Keep the handoff behavior stable when length is the factor you want to compare. A shorter page that also opens a different product route or spawns another tab changes more than its reading burden.

Write one route record for each variant: entry source, page version, offer, product destination, opening mode and outcome definition. Then identify the intended difference. An observed result remains a comparison of those complete routes; it does not become an effect of word count just because one version contains fewer words.

| Difference between variants | What the comparison investigates | What it cannot isolate |

|---|---|---|

| Length changes; offer and navigation stay fixed | The planned content-length treatment | Every individual section’s contribution |

| Tab behavior changes; content stays fixed | A navigation-context hypothesis | A benefit caused by shorter copy |

| Length, destination and tab behavior all change | A combined route change | Which of those factors explains the result |

NN/g’s research on new tabs emphasizes the reader’s task and device rather than a universal rule based on destination type. Its usability observations do not establish a conversion lift for a particular presell route.6

Use the same-tab versus new-tab worksheet before setting up the comparison. Record uncertainties and plan the sample and stopping rule separately. A page left open in a background tab is not evidence that its argument persuaded the reader.

English explanation of navigation links and local actions. Use it to inspect the handoff’s semantics; it does not compare long versus short copy or establish a conversion result. Video: Buttons vs. Links: What’s the Difference and Why Does it Matter?, by NNgroup, on YouTube. It loads from YouTube only when you press play.

How long should your page be?

Interactive

Find the right length for your next page

Five questions about your reader, your product and your traffic.

  1. When they click, do readers know they have the problem?
  2. How does the price feel to a first-time buyer?
  3. Does your product work in a way people have never heard of?
  4. How many serious doubts do buyers raise before paying?
  5. Where does the traffic come from?

Before you decide the length

7 items

Equal scroll percentages can refer to different explanations

When comparing a shorter and a longer presell, map each threshold to the content in that exact version. Cutting a section changes the relationship between page height and the argument. An equal percentage label can now refer to a different paragraph or CTA, so it should not automatically be called equal exposure to the same message.

Google Analytics documents a specific enhanced-measurement scroll condition rather than a comprehension test.7 The scroll depth tracking guide explains why a reached position needs its event definition and page version. Keep a missing threshold record separate from a conclusion that the page was too long. Other limitations in the setup may remain unresolved.

For a fictional long explanation of a cup’s construction, record which question the removed section answered. Compare the changed reading path and intended next step rather than trying to maximize a depth percentage. If shortening the page also changes the offer or product destination, document those interventions separately. Preserve the earlier version and the conditions of the comparison. Depth can help choose the next inspection; it cannot by itself decide which length sells better or supply an order result.

RecordUseful reviewDo not infer
Version lengthMap position to contentEqual percentage means equal reading
Section removedName its explanatory jobShorter automatically sells better
Observed thresholdKeep definition and setupComprehension or completed orders

When a long page contains distinct reader tasks, a shortcut can be more useful than removing the answer. The presell jump-link guide shows how to check labels, section targets and keyboard flow.

Frequently asked questions

Is long-form copy dead?

No. In our research database, presell advertisers write longer ads than stores do, and nearly one presell ad in five runs past 250 words. What does not work is padding: paragraphs that repeat, introductions that delay, adjectives instead of facts.

How many words should a presell page have?

There is no right number. The nine example pages in this article run from about 550 words (a quiz landing, before the personal result) to about 2,500 (an advertorial). Decide by what the reader must learn, then cut everything that does not help them learn it.

Do long pages work on mobile?

Yes, when they are easy to skim: short paragraphs, subheads that carry the argument, images that prove a point, and buttons placed after the moments of conviction. Write and review the page on a phone first, at the size your readers will see it.

Should a long page have only one button, at the end?

No. Put a button after the product is introduced, after the proof, and at the end. Readers who are convinced early should not have to scroll through the rest to act.

Does an audio version help a long page?

It gives readers who would not read 2,000 words another way in. Several of our example pages offer the article as audio at the top. It does not replace a skimmable page: most visitors will still scan the text.

Sources

  1. Ad copy length in our research database (1,287,551 ads), Ads collected from the public Meta Ad Library. Ads started April 11 to October 11, 2026
  2. How Little Do Users Read?, Jakob Nielsen, Nielsen Norman Group (2008)
  3. F-Shaped Pattern of Reading on the Web: Misunderstood, But Still Relevant (Even on Mobile), Kara Pernice, Nielsen Norman Group (2017)
  4. How Long Do Users Stay on Web Pages?, Jakob Nielsen, Nielsen Norman Group (2011)
  5. Breakthrough Advertising, Eugene M. Schwartz (1966)
  6. Opening Links in New Browser Windows and Tabs, Jakob Nielsen and Anna Kaley, Nielsen Norman Group (September 27, 2020; accessed October 26, 2026)
  7. Enhanced measurement events, Primary documentation

ClickBanger Editorial We build presell pages for e-commerce brands and spend our days reading ads, landers and test results. Questions or a correction? Write to us.

Keep reading

All articles

Testing & growth13 min read

A/B testing presell pages: read the result clearly

Read a presell A/B test against the plan you wrote before launch. Keep the same visitor basis, check the product handoff, and separate clicks from purchases. A result below your evidence threshold remains uncertain; it does not prove that the pages perform alike.

From product URL to presell page in minutes

Advertorial, listicle, quiz, comparison and 12 more formats, written from your product page and ready to edit.