What you'll understand by the end of this lesson
- Why visitors miss information that's clearly visible on the page
- What the F-pattern tells you about how people scan web pages
- Why putting key content in unexpected places hurts conversion
- How to design for multiple mental states on the same page
The principle in plain English
The human brain can't process everything it sees. To cope with the information overload of everyday life, it applies a filter: pay attention to what's relevant to the current goal, ignore everything else.
This is selective attention. It's not a flaw — it's an efficient system that works well most of the time.
On a website, it means that a visitor arriving with a specific goal — finding the price, checking a return policy, reading a feature list — will scan the page for signals that match that goal. Content that doesn't pattern-match to what they're looking for gets filtered out, even if it's large, bold, or colourful.
This is why "but it's right there on the page" is not a defence when users miss information. If it wasn't in the place they expected to look, they didn't look there.
A simple example
Think about the last time you walked into a supermarket looking for milk. You didn't consciously read every shelf label. You filtered for the dairy aisle — possibly using a previous mental map, possibly using visual cues like fridges and packaging colours. Everything else was background.
If the supermarket moved milk to aisle 3 instead of the back corner, you'd probably miss it on the first visit — even if there was a large sign. Your attention was filtering for the expected pattern, not scanning everything equally.
The same thing happens on every web page you've ever built.
The F-pattern
Eye-tracking research from the Nielsen Norman Group showed that users reading web pages follow a broadly F-shaped pattern:
- A horizontal scan across the top of the page
- A second, shorter horizontal scan slightly lower
- A vertical scan down the left side
The right side of the page — especially the lower-right — receives very little attention.
This doesn't mean every user reads in a perfect F. It means the distribution of attention is not equal. The top-left is prime real estate. The bottom-right is a dead zone.
The practical implication: if your most important CTA, your key differentiator, or your trust signal sits in the lower-right of a content block, a large share of visitors will never see it. Not because they left — but because their attention never went there.
Check where your primary CTA sits relative to the F-pattern. If it's only visible after a vertical scroll, or positioned to the right of a dense text block, it's in attention shadow. Move the key action to where the eye naturally lands — typically the left side, in the top half of the content above the fold.
Designing for multiple mental states
One of the harder CRO challenges is that the same landing page serves visitors in very different states.
Some visitors are in research mode — they arrived to evaluate, not buy. They're reading carefully, comparing, looking for reasons to trust. They'll read your case studies and your about page.
Others are in decision mode — they've already done their research elsewhere and arrived ready to act. They're scanning for the price and the buy button. Everything else is friction.
If your page is designed only for the research-mode visitor, the decision-mode visitor can't find the CTA quickly and leaves. If it's designed only for the decision-mode visitor, the research-mode visitor doesn't get enough information to build trust and also leaves.
The solution isn't to choose one. It's to design a hierarchy that serves both:
- The decision-mode path: clear headline → key value → CTA, all visible above the fold
- The research-mode path: supporting content, FAQs, testimonials, and case studies lower on the page
The decision-mode visitor gets what they need quickly. The research-mode visitor scrolls to find what they need. Neither has to fight the layout.
Why unexpected placement kills conversion
Design teams sometimes place CTAs in unusual spots to be distinctive or to follow a visual design concept. A button buried in the middle of a paragraph. A "get started" link in the page footer rather than the hero.
The problem is that selective attention is habit-forming. Users expect interactive elements — buttons, links, forms — to appear in conventional positions: top-right navigation, end of a hero section, bottom of a pricing card. When they're somewhere else, the attention filter misses them.
Being different with CTA placement doesn't create delight. It creates confusion. Conversion drops not because the user doesn't want to act, but because they couldn't find where to.
Unconventional page layouts — no header, floating elements, non-linear scrolling — can look beautiful in a design review. They often perform badly in conversion because they break the mental map users arrive with. If your layout requires the user to learn a new navigation model before they can do what they came to do, you've added friction before they've even engaged with your content.
The CRO audit
Walk through your key landing pages and ask:
1. Is your primary CTA in the top-left zone of the F-pattern?
Your most important action should sit where attention naturally lands. Above the fold, left-aligned or centred, with enough contrast to be seen during a quick scan — not after careful reading.
2. Can a visitor in decision mode find the price and the CTA in under five seconds?
Time yourself. If you can't find them quickly knowing where they are, a first-time visitor in a hurry has almost no chance.
3. Are there elements on the page that only work if the user reads carefully?
Key information buried in body copy, trust signals in footnotes, guarantees in the small print — if it requires focused reading to find, most users won't find it.
A product page has a compelling guarantee — 'Full refund, no questions asked, within 60 days' — written in the fifth paragraph of the body copy. Very few users seem to notice it. What's happening?
You've seen how users filter what they look at. But there's another bias that distorts what CRO teams learn from their data — one that makes winning tests look more reliable than they are.