Visual semantics

What Are Macro Context and Micro Context?

Macro context is the main content of a page, placed above the fold, and micro context is the supplementary content below it. Macro context carries the main predicate and the core answer or function. Micro context carries secondary attributes, related entities, and most internal links.

Macro and micro context, illustrated
A page stacked top to bottom. H1, standfirst, main component, and primary call to action form the macro context. Secondary attributes, related entities, FAQs, and the identity block form the micro context
The top three blocks are macro context and answer the query. Everything below widens the page to its neighbors and carries the links.

What are macro context and micro context on a web page?

Macro context is the page's main content and micro context is its supplementary content, the same split Google's rater guidelines draw.

Google's Search Quality Rater Guidelines separate Main Content from Supplementary Content. Main Content directly helps the page achieve its purpose. Supplementary Content supports the visit without being the purpose. Macro and micro context are the build terms for the same two regions.

Centerpiece annotation decides one block. The macro and micro split decides every block on the page. The centerpiece sits inside the macro context as its first block after the H1. Every other block is assigned to one side of the line.

The split is a meaning split, not a styling choice. One predicate runs through the macro context. Micro context introduces the neighbors: other attributes, related entities, and next steps. Mixing the two blurs what the page is about. A page that argues two predicates above the fold reads as two pages sharing one URL.

What belongs in the macro context above the fold?

The macro context holds four blocks: the H1, the standfirst, the centerpiece component, and the primary call to action.

Most of a page's semantic weight sits in the macro context. The query's answer and the main predicate live here in full. A reader who stops at the fold already has the answer.

Nothing interrupts those four blocks. Ads, newsletter signups, and popups stay out of the space between the H1 and the centerpiece. A secondary attribute that needs more than one paragraph moves below the fold. Micro topics that leak upward thin the main predicate. Chat widgets and promo strips stay outside the main element.

Take a project management SaaS feature page for task dependencies. The macro context names the feature, states what it does, shows the dependency view, and offers the trial. Integrations, pricing tiers, and customer stories wait below. Each one is useful, and none of them is the answer to the query. The trial button sits beside the view, not three screens down.

What belongs in the micro context below the fold?

The micro context holds secondary attributes, related entities, FAQs, experience blocks, and most of the page's internal links.

Micro context carries the rest of the page's meaning. The page widens here from the query to its neighbors. The drill-downs and next steps a generative model produces for the query find their blocks in this region. A next step with no block on the page and no page in the cluster sends the session elsewhere.

Tabbed selectors compress several sub-entities into one component. Each tab holds a short definitional snippet that mirrors the opening of the child page, plus a link to it. The tab content sits in the DOM on load, not behind a click. The same tab taxonomy repeats in the FAQ block, which keeps the context stable.

Conversion bands belong here too. A band follows each trust escalation: after the definition, after the service list, and before the FAQs. Objection handling sits inside the band as text. The primary call to action stays in the macro context, and the bands repeat it where a reader is ready. Each band carries one action, never a menu of options.

Internal links sit in the micro context, from the middle to the bottom of the page, and never inside the centerpiece.

Link density is low above the fold and high below it. A wall of links near the top reads as navigation or an aggregator, not as main content. The centerpiece carries no internal links at all. Links earn their place after the answer, where a reader looks for the next step.

Each target page receives one canonical anchor phrase, used on every link across the site. Variation happens only inside the dual-entity form of that phrase. A link to the same target appears once per page, not in every section. Card grids that link to child pages use each child's H1 as the anchor. Anchors never read click here or read more.

A lapsed Google patent on query augmentation treats a click on another link on the landing page as a long click. The same patent stores anchor text repeated across many pages as a query the target answers. The patent has expired, so this is design logic, not a guarantee. Links placed where a satisfied reader looks next earn the click, and consistent anchors name the target.

How do Google's segmentation patents draw the line between blocks?

Google's segmentation patents split a page into blocks by white space and spatial boundaries, then read the blocks in order.

A granted Google patent on document segmentation based on visual gaps divides a document by its white space into a hierarchy of blocks. Large gaps separate major regions. Smaller gaps separate the blocks inside them. The line between macro and micro context is one of those gaps, placed on purpose. Designers set those gaps through spacing, borders, and background changes.

Google's patent on layout-aware multimodal pretraining defines a block as a part of the document with a clear spatial boundary to its neighbors. The encoder sorts blocks by their upper-left corner, vertical position first. The order becomes part of the stored document representation. Blocks without a clear boundary read as one block.

The Content Warehouse API leak shows Google labeling a main flow of text and separate flows for boxes and asides. Side boxes and callouts are therefore built as separate blocks, not as inline interruptions. One uninterrupted main flow keeps the macro context readable as one unit.

Sources

Patents and filings describe systems Google has invented. None is a confirmed ranking factor.

What is centerpiece annotationCenterpiece annotation is Google's label for the primary content block of a page, extracted from its HTML structure.Visual semanticsVisual semantics is the meaning a search engine reads from a page's layout, blocks, and components alongside its text.What is a topical mapA topical map is the plan of every entity, attribute, and query a site covers.Query Deserves a PageQuery Deserves a Page is the test that decides if a query variation gets its own page, a section, or a sentence.Website DevelopmentFive platforms, chosen by what the architecture needs, not by preference.Semantic SEO ArchitectureSeven documents with a fixed structure. Every engagement produces all seven.Talk to us speech bubble: message the Digital Vikingz team on WhatsAppTalk to usMessage the team on WhatsApp

Questions about macro and micro context

Is the fold a fixed line in pixels?
No. Screen size moves the fold on every device. The split is defined by block order instead: H1, standfirst, main component, and primary call to action come first. A small screen pushes the fold higher without changing that order.
Can reviews and ratings sit in the macro context?
Yes, on high-risk and high-cost decisions. A short evidence strip with credentials and a review rating sits before the definitional block on those pages. Low-risk pages put the definition first. Full testimonial blocks never displace the main component.
Does a sidebar beside the main content count as macro context?
Yes, by position. Blocks at the same height as the centerpiece are serialized directly before or after it. A sidebar at that height holds only entity-relevant content, such as key facts or a spec summary. Ads, promos, and link walls never sit beside the centerpiece.
Are the site header, menu, and footer part of the micro context?
No. The header, menu, and footer are site furniture shared by every page. Micro context is the page's own supplementary blocks: secondary attributes, related entities, and FAQs. Both sit outside the centerpiece, but only micro context is written for this page.
Where does the buying guide go on an ecommerce category page?
Below the product grid, in the micro context. The filtered grid is the category page's main component and sits first. The guide answers the questions a shopper asks after scanning the products.
When does a micro-context block become its own page?
A block becomes a page once its query passes Query Deserves a Page on demand, entity, similarity, predicate, and pattern. Until then it stays a section with its own id and opening sentence. The section's id becomes the new URL when the split happens.
Where do links to external sources sit?
External links sit in their own block, separate from on-site navigation and internal links. Google's segmentation work distinguishes off-site link blocks from on-site ones. A sources list at the end of the micro context keeps both clean.
Does the macro and micro split apply to blog posts?
Yes. The standfirst and the first sections answer the post's query in the macro context. Related posts, the author bio, and tag lists sit in the micro context. A post with related links above its first paragraph opens with navigation instead of an answer.
Does the split change between mobile and desktop?
The order changes when a sidebar stacks above or below the main column on a small screen. The check runs on both renders. The main column keeps the same sequence on both, and stacked blocks land after the centerpiece.
How does the split work on a homepage?
The macro context declares the central entity and proves it: the H1, a definitional block, and evidence such as credentials or reviews. The micro context holds the service set, usually as a tabbed selector or a card grid. Nothing about a single sub-service appears before the definitional block.
How is the macro and micro split checked on an existing page?
Mark the end of the primary call to action and list every block above that mark. Anything other than the H1, standfirst, main component, and call to action is a leak. Then count the internal links above the mark, which ideally number zero outside navigation.
Usman Ishaq, founder and CEO of Digital Vikingz
Written by Usman IshaqFounder and CEO, Digital Vikingz. Koray Tuğberk Gübür Semantic SEO Fundamentals certified. Mechanical engineering background.Reviewed September 2026. Profile
Book a callWhatsApp