Visitors spend 57% of their viewing time in the first screenful of a page and 74% in the first two, according to Nielsen Norman Group's 2018 eyetracking study of 120 participants and more than 130,000 fixations. That number decides most of what follows: landing page structure is the job of putting the offer, the proof and the action where the attention is, and making everything else earn a scroll.
This guide sets out the nine blocks we build into a B2B landing page, in the order a visitor meets them, and the evidence for why each sits where it does. The numbers come from Nielsen Norman Group, Unbounce's 2024 Conversion Benchmark Report, Zuko's 2025 form benchmarks, Google's own documentation, Portent and the Spiegel Research Center, with vendor figures labelled as such.
Why the order matters more than the sections
Most landing page advice lists the same ingredients. Unbounce's own anatomy guide, a vendor description, names a unique selling proposition, a hero image or video, benefits, social proof and a single conversion goal. The list is fine. It just does not say which block gets the 57% of attention and which gets the long tail.
Nielsen Norman Group's 2018 study is the clearest published picture of where attention goes. Its 2010 study found 80% of viewing time above the fold; by 2018 that had fallen to 57%, but the shape held.
| Screenful | Share of viewing time | Blocks that live there | What they have to achieve |
|---|---|---|---|
| First | 57% | Hero: headline, subheading, primary CTA, visual | Say what it is, who it is for, and what to do next |
| Second | 17% | Trust strip, problem statement, first benefits | Make the claim believable and relevant |
| Third and below | 26% in total | How it works, social proof, objections, form, final CTA | Answer the questions of people still reading |
Source: Nielsen Norman Group, Scrolling and Attention, 2018 eyetracking study of 120 participants
The rule that falls out of this: the first screenful has to work as a complete argument, because for most visitors it is the only one they see. The lower blocks are written for the minority who scroll, and that minority is the one that converts.
The landing page structure, block by block
This is the order we use on paid-campaign and demo pages. It is a default, not a law.
| Block | What it holds | Its one job | Evidence it leans on |
|---|---|---|---|
| 1. Hero | Headline, one-line subheading, primary CTA, one visual | Finish the ad's sentence and name the action | NN/g 57% above the fold; Google Ads message match |
| 2. Trust strip | Four to six client logos or one hard number | Make the headline believable in a glance | NN/g first impressions, 50 ms judgement |
| 3. Problem | Two or three sentences on the visitor's situation | Confirm the page is for them | NN/g scanning data, 79% scan |
| 4. Benefits | Three or four outcomes, each a heading and one line | Translate features into results | Unbounce 2024 readability correlations |
| 5. How it works | Three steps or three product screens | Remove the "but how" objection | Google Ads: useful, unique content |
| 6. Social proof | One to three quotes with name, role, company, result | Let a peer make the claim | Spiegel 2017, plateau after five reviews |
| 7. Objections | Three to five short answers to common sales questions | Clear the last doubts before the form | Google Ads: key information near the top |
| 8. Form | The fewest fields sales will accept, single column | Convert the visitor who scrolled this far | Zuko 2025; NN/g 2016 form guidance |
| 9. Final CTA | The headline promise and the button again, minimal footer | Catch the reader who finished the page | NN/g 2018, long tail of attention |
Source: Octo Agency build order; evidence as named in each row and in the Sources list
Block 1, the hero. The headline restates the ad's promise, the subheading gives the one fact that makes it plausible, and the button says what happens next in the visitor's words ("Book a 20-minute demo", not "Submit"). Google's Ads guidance says the page should mirror the ad's call to action: an ad offering a free tour leads to the sign-up form, not a homepage. Our PPC post covers message match in detail.
Block 2, the trust strip. Nielsen Norman Group cites research showing aesthetic judgements form within about 50 milliseconds of arriving and rarely change afterwards. A row of recognisable logos, or one plain number, lands inside that window.
Block 3, the problem. The block most often cut, and the one that most often earns its place with colder traffic. Two or three sentences that describe the visitor's situation do more for relevance than another benefit. Keep it short: Unbounce's 2024 report, drawn from more than 41,000 landing pages and 57 million conversions, found word count correlated negatively with conversion at -18.6%.
Block 4, the benefits. Three or four, each with a heading that stands alone, because 79% of users scan. Each is an outcome with one line of how. Nielsen Norman Group's early writing study, old but never contradicted, found concise text improved measured usability by 58%, a scannable layout by 47%, objective language by 27%, and all three together by 124%.
Block 5, how it works. Three steps or three screens. Buyers who scroll this far are asking what it would look like for them, and a screenshot answers faster than a paragraph. Google Ads asks for original content that is clear about what the product is and does.
Block 6, social proof. One to three testimonials with a name, a role, a company and a result. The Spiegel Research Center found the lift from reviews levels off rapidly after the first five, and that purchase likelihood peaks between 4.0 and 4.7 stars and falls toward a perfect 5.0. Three specific quotes beat twelve vague ones.
Block 7, objections. Pricing, security, implementation time, integrations, contract length: whatever sales hears on every first call. Three to five short answers written as statements, not an accordion, because a hidden answer is an unanswered one.
Block 8, the form. Where the page converts or does not. It has its own section below.
Block 9, the final CTA. Repeat the promise and the button. The footer carries legal links and nothing else; a full site navigation here is an exit door at the moment the reader has decided.
The hero carries the page, so it carries the readability rules
The Unbounce correlations bite hardest in the hero, because that is where every visitor reads.
| Measure | Correlation with conversion rate | What it means for the hero |
|---|---|---|
| Difficult words (three or more syllables) | -24.3% | Replace "comprehensive" and "infrastructure" with short words |
| Reading time | -19.4% | One headline, one subheading, one button; no paragraph |
| Word count | -18.6% | Cut the second sentence |
| Reading level: 5th to 7th grade | 11.1% conversion | The target for the headline and benefits |
| Reading level: professional | 5.3% conversion | Where most B2B hero copy sits today |
Source: Unbounce Conversion Benchmark Report 2024, 41,000+ landing pages, 464 million visitors, 57 million conversions; vendor data
The difficult-word correlation was -15% in Unbounce's 2020 report, so the 2024 figure is 62% stronger. Unbounce notes one exception, financial services, where more complex language can help. For everyone else the reading level of the hero is a structural decision, because it decides whether the 57% of attention becomes a click.
Device is the other constraint. Unbounce's sample was 83% mobile visits, with desktop converting 8% better. On a phone the first screenful is the headline, two lines of subheading and a button. Build the hero for that viewport first.
The form block: the fewest fields sales will sign off
Zuko's 2025 benchmark covers 93 million form sessions. Enquiry forms, where most B2B landing pages sit, complete at 42.7% and carry an average of 17.4 fields.
| Form type | Sessions tracked | Completion rate | Desktop | Mobile | Average fields |
|---|---|---|---|---|---|
| All forms | 93,022,997 | 51.71% | 54.48% | 47.53% | Not stated |
| Enquiry (lead generation) | 4,797,087 | 42.7% | 47.7% | 37.7% | 17.4 |
| Contact | 321,937 | 47.4% | 51.2% | 39.4% | 19.2 |
Source: Zuko form analytics benchmarks, 2025; vendor data from a form-analytics company
Seventeen fields on a form that converts under half its starters is the clearest structural leak on most pages. Nielsen Norman Group's form guidance, from 2016 and still current, is to remove every field that can be derived elsewhere, collected later or dropped; keep a single column, because multiple columns interrupt the vertical momentum of moving down the form; and cut optional fields or clearly label the one or two that remain. NN/g cites a 2014 study in which forms following usability guidelines were submitted error-free first time 78% of the time, against 42% for forms that broke them.
The structural decision is who owns the field list. If sales owns it, the form grows; if marketing owns it, sales complains about lead quality. The compromise that holds is a short form (name, work email, company, one qualifying question) with enrichment and routing after the submit. Our post on form conversion statistics has the field-by-field data.
Speed and Quality Score sit underneath every block
Structure is what the visitor sees. Two things they do not see decide whether they see it at all.
Speed. Portent's 2022 study of 14 B2B lead-generation sites and 5.6 million sessions found conversion rates fall steeply from one to three seconds of load time and then level off; a site loading in one second converted at three times the rate of one loading in five, and five times the rate of one loading in ten. Google's Core Web Vitals thresholds, measured at the 75th percentile of page loads, are a Largest Contentful Paint of 2.5 seconds or less, Interaction to Next Paint of 200 milliseconds or less and Cumulative Layout Shift of 0.1 or less. For the hero that means one optimised image or a poster frame, not an autoplaying video.
Quality Score. Google Ads rates landing page experience as one of three components, with expected click-through rate and ad relevance, each scored above average, average or below average against advertisers who showed for the same searches in the last 90 days. Google's guidance maps onto the nine blocks: match the ad and keywords, put key information near the top so people do not have to scroll for it, keep navigation simple, provide useful and unique content including real customer reviews, and make the page quick and usable on a phone.
Neither changes the order of the blocks. Both change what each block may contain.
When to cut, merge or reorder
The nine blocks are a default for a cold or warm paid-search visitor. Three variables change the order.
Traffic source. Unbounce's 2024 data puts email at a 19.3% median conversion, paid social at 12% and paid search at 10.9%. An email visitor already knows the sender, so the trust and problem blocks can shrink or go and the form can move up to the second screenful. Our benchmark post breaks down conversion rates by industry and traffic source.
Offer weight. A gated PDF needs blocks 1, 2, 8 and 9 and little else. A demo request for a six-figure contract needs every block, and often a second proof block beside the form. Ask of each block whether the visitor needs its answer before they will act.
Attention decay. Contentsquare's 2026 benchmark across 6,500 sites and 99 billion sessions found engagement down 10% year on year and conversion down 5.1%. The answer is not a shorter page for everyone; it is a hero that converts on its own and lower blocks that each repeat the CTA, so a reader who stops at block 5 still has a button in view.
What we do not recommend is the inverted order some templates use, with the form in the hero and the argument below. For a known brand on email traffic it works. For a search visitor who has never heard of the company, it asks for a commitment before the case is made.
What this means for your next landing page
Build the hero as if it were the whole page, because for 57% of viewing time it is. Write it at a reading level a twelve-year-old could follow, with one headline, one subheading and one button that names the action. Put the trust strip under it and the problem under that, then benefits, product, proof and objections, each with a heading that carries its point to a scanning reader. Make the form the shortest one sales will sign, in a single column, and repeat the CTA at the end. Then check load time, because a page that does all of this and paints in five seconds is still losing two thirds of its conversions before anyone reads the headline.
We build landing pages this way inside our four-week marketing-site process, in Webflow, with the block order agreed before design starts. Our marketing websites service page explains how we scope it.


