Web Design

Layout Archetypes — F-Pattern, Z-Pattern, Bento

Layout Is a Content Decision, Not a Style Decision

The layout of a web page is not a visual preference — it is an answer to two questions: what is the nature of this content, and what does the user need to do with it?

A bento grid is right for a visual portfolio because the user’s task is browse and be impressed. It is wrong for a clinic explaining its treatment process because the user’s task is read and evaluate in order. Apply bento to the clinic and the layout actively impedes the user’s task, regardless of how good it looks in a Figma file.

The most expensive web design mistake is choosing a layout because it was trending on Dribbble last month. You pay for it in bounce rates, lower time-on-page, and conversion suppression — none of which the client immediately connects to the layout decision.

Here is the practitioner’s guide to matching layout to content.


F-Pattern — For Text-Heavy, Information-Dense Pages

What it is

In the late 2000s, the Nielsen Norman Group conducted eye-tracking research on how users read web pages with significant text density. The pattern that emerged was shaped like the letter F:

  • First horizontal stroke: users read the top line nearly fully, left to right.
  • Second horizontal stroke: users read a second line partially — shorter than the first.
  • Vertical stroke: users scan down the left margin, picking up the first word or two of each line.

Most content never gets read past the left margin for the majority of users. This is not a flaw in your users — it is rational information-foraging behaviour.

Works best for

Blog posts and long-form articles. Product comparison pages. Directory listings. Search results. Documentation pages. Any context where users are extracting specific information from a text-heavy source.

How to design for it

Put your most important content at the F’s entry points:

  • Top-left: the most critical headline or value signal
  • First line across: the lead sentence or key claim, because it gets the most complete read
  • Left margin: the first word of each paragraph and heading, because this is the scanning path

Left-align all body text. Use clear, weight-differentiated headings so users can anchor to them during the left-margin scan. Keep paragraphs short — three to five lines maximum. Use bullet lists and sub-headings to create left-margin anchor points.

What breaks it

Centre-aligned text on information-dense pages. Equal-weight headings that give the eye no anchors. Walls of text without sub-headings. Decorative left-margin elements that compete with the scanning path.


Z-Pattern — For Marketing and Landing Pages

What it is

On pages with low text density and visual anchors — marketing pages, landing pages, home pages — eye-tracking shows a different pattern: users move top-left → top-right → diagonal down-left → bottom-right. The shape is a Z.

This happens because low-density pages don’t trigger line-by-line reading. The eye moves from anchor to anchor, following the largest, most distinct visual elements.

Works best for

Landing pages. Home pages. Service pages. Campaign pages. Any page where the primary goal is a single conversion action and the content is visual-first.

How to design for it

Map your most important elements to the Z’s four anchor points:

  • Top-left: logo or brand mark (trust anchor)
  • Top-right: navigation or a secondary CTA (“Sign in”, “Contact”)
  • Center-left: primary value proposition headline
  • Center-right: hero image or supporting visual
  • Bottom-right: primary CTA button

The Z’s exit point — bottom-right — is where conversion happens. Your most important action should sit there.

What breaks it

Too much content density on what was intended to be a Z-pattern page. The moment a page has enough text that users start reading linearly, it reverts to F-pattern behaviour — but without the design structure to support F-reading, it fails at both.


Single Column — The Underrated Layout

What it is

One continuous content column, typically constrained to 640–800px maximum width for reading optimisation, centred on the page with generous side margins on desktop. The user moves in one direction: down.

Works best for

Mobile-first experiences. Long-form storytelling and editorial content. Email design. Checkout flows. Onboarding sequences. Case studies. Any experience where the user must complete a sequential task or read content in full.

Why it works

Single-column removes all left-right navigation decisions. The user cannot scan horizontally to find something more interesting — they move through the content in the order you intended. This reduces friction for task completion and ensures sequential content is consumed in sequence.

For checkout flows specifically, this is not a limitation — it is a feature. Every unnecessary visual choice you remove from a checkout page statistically reduces abandonment.

The mistake designers make

Treating single-column as simple, lazy, or low-effort. It is the opposite. Single-column is the most unforgiving layout for poor content hierarchy — there is nowhere to hide bad prioritisation. If your most important point is buried in paragraph four, a three-column layout at least gives users something else to look at. Single-column forces your content to earn attention in the order you arranged it.

Getting single-column right requires sharper copywriting, more deliberate heading hierarchy, and stronger visual pacing than almost any other layout.


Bento Grid — The Portfolio and Dashboard Layout

What it is

An asymmetric grid of cards at varying sizes — named after Japanese bento boxes, where different foods sit in different-sized compartments. Popularised by Apple’s product pages in the early 2020s and adopted widely in creative and tech design.

The defining feature: card size communicates importance. The biggest card gets the most visual weight, the most attention, and the first read.

Works best for

Creative portfolios. Product feature showcases. SaaS dashboards. Personal site hero sections. Any context where content is visual-first, items are not directly comparable, and you want hierarchy through physical size rather than position alone.

Why it works

Asymmetric grids create visual interest through contrast — variation in scale is inherently more engaging than uniform repetition. The largest card naturally becomes the anchor point from which the eye moves to explore the rest. Users don’t read bento grids linearly; they browse.

When it fails

Service businesses listing comparable offerings — the size variation implies some services are more important than others, which is usually not the intended message.

Content requiring logical linear flow — a process, a timeline, a sequential explanation.

Any context where users need to compare items side by side — asymmetric cards make comparison significantly harder than uniform card grids.

The trend problem

Bento grid appears in roughly half of all Dribbble and Behance UI concept posts. It looks exceptional in mockups. It is frequently wrong for the actual content it gets applied to — clinic service listings, restaurant menus, law firm practice areas. When the content is uniform and text-dominant, bento imposes a false hierarchy and makes the layout actively harder to parse.

Choose bento when visual scale difference genuinely represents content importance difference. When it doesn’t, you are optimising for portfolio aesthetic at the expense of user experience.


Card Grid — For Parallel, Comparable Content

What it is

A uniform grid of equal-size cards — typically two to four columns — where each card represents one item in a comparable set.

Works best for

E-commerce product listings. Article and blog index pages (when thumbnail-browsable). Team member profiles. Service packages (when users need to compare across options). Directory entries.

The card grid trap

Card grids fail when the content inside the cards is not genuinely parallel. Mixing articles with very different text lengths, different metadata, or different visual weights breaks the grid’s uniformity and creates awkward forced-height situations. Forcing all cards to the same height with truncated text or excessive white space damages scannability.

Before using a card grid, ask: is each item in this set genuinely comparable to the others, and will users want to scan across items to choose? If yes, card grid. If no, reconsider.


Decision Table — Choosing Your Layout

LayoutBest Content TypeUser TaskWorst Misuse
F-PatternText-heavy, information-denseExtract specific informationMarketing pages — wastes F’s scanning efficiency
Z-PatternLow-density, visual-anchor pagesAct on a single CTAContent-heavy pages — reverts to broken F-reading
Single ColumnSequential, narrative, task-flowRead fully or complete a taskVisual portfolio — kills browsing; forces linear read
Bento GridVisual, asymmetric, portfolio-styleBrowse and be impressedUniform service listings — creates false hierarchy
Card GridParallel, comparable itemsScan and compareNon-parallel content — breaks at different text lengths

Not sure which layout fits your project? Free consultation →


References

  1. Nielsen Norman Group — F-Shaped Pattern of Reading on the Web. https://www.nngroup.com/articles/f-shaped-pattern-reading-web-content/
  2. Nielsen Norman Group — Eyetracking Web Usability. https://www.nngroup.com/books/eyetracking-web-usability/
  3. Baymard Institute — Checkout Usability Research. https://baymard.com/research/checkout-usability
  4. MDN Web Docs — CSS Grid Layout. https://developer.mozilla.org/en-US/docs/Web/CSS/Guides/Grid_layout
  5. Apple — Mac product page bento grid design reference. https://www.apple.com/mac/
  6. Interaction Design Foundation — Visual Hierarchy. https://ixdf.org/literature/topics/visual-hierarchy

Web Design Layouts — Common Questions

Can I use a bento grid layout for a service business website?

Only if your services are genuinely visual and portfolio-driven. A bento grid creates hierarchy through card size — the biggest card gets the most attention. If your services are roughly equal in importance and need descriptive text to explain them, a card grid or a single-column service list will outperform bento every time. Bento grids are frequently applied to service businesses because they look impressive in mockups, but they make comparison and scanning significantly harder for users trying to evaluate options.

What layout should I use for a landing page that needs to convert?

Z-pattern for a short, high-impact landing page with a single CTA. The Z traces the natural reading path on low-density visual pages: logo top-left, nav top-right, value proposition center-left, supporting visual center-right, CTA bottom-right. Match the Z's exit point to your conversion action. If the page is longer with multiple sections of information, stack Z-patterns vertically — each section follows its own Z trajectory.

Is single-column layout really suitable for desktop, or is it just for mobile?

Single-column is underused on desktop and frequently more effective than designers expect. Long-form content, case studies, checkout flows, and onboarding sequences perform well in single-column on desktop because they eliminate left-right decision fatigue. The max-width constraint — typically 680–800px for reading-optimised text — combined with generous white space on both sides is not a design limitation; it's a reading experience enhancement. Many high-conversion SaaS checkout pages use single-column precisely for this reason.

How do I choose between a card grid and an F-pattern layout for an article listing page?

It depends on whether visual thumbnails are the primary differentiator. If users browse by image (photography, portfolio, visual products), card grid wins — the thumbnails drive engagement. If users scan by title, category, or date — which is how most people browse blog and news content — F-pattern wins. Design the listing for the actual scanning behaviour: left-aligned titles, visible metadata, and no center-aligned text. Wasting the left-margin entry point of the F-pattern with decorative elements is one of the most common content-listing layout mistakes.