Internal links connect pages within the same website, while external links send visitors to pages on another website. That destination difference changes how each link functions. Internal linking builds navigation paths, connects related content, and helps search engines find pages. External linking directs readers towards independent evidence, official information, useful tools, or supporting resources.
Both can provide contextual signals through anchor text and surrounding copy, but neither guarantees better rankings. Effective linking depends on destination relevance, technical accessibility, editorial purpose, site structure, and regular maintenance.
Internal, External, and Inbound Links Are Different
Three link types often appear in SEO discussions, but they describe different relationships.
An internal link points from one page to another page within the same website context. A product category linking to a product page provides a straightforward example.
An external or outbound link starts on the website being viewed and points towards another website. A business resource linking to official government documentation creates an outbound link from the business website.
A backlink, also called an inbound link, works from the opposite perspective. Another website links to the website being evaluated. Therefore, an outbound link from Site A to Site B becomes an inbound backlink for Site B.
Keeping these terms separate prevents faulty audits. Internal-link analysis examines relationships within a site. Outbound-link analysis examines destinations that a site references elsewhere. Backlink analysis examines links arriving from other websites.
Each category requires different maintenance, risk assessment, and strategic decisions.
How Internal Links Work Across a Website
Internal links give users routes between related pages while creating connections that search crawlers can follow. Their usefulness extends beyond adding clickable text to content.
A service page may link to a detailed resource explaining a technical issue. That resource may direct readers towards another supporting page or a relevant enquiry pathway. Meanwhile, categories can connect products, subcategories, buying resources, and other useful destinations.
These relationships gradually form website architecture.
Internal links can also indicate context. Descriptive anchor text and surrounding copy help communicate what users should expect at the destination. However, placing a link does not automatically improve the destination’s rankings. Search systems consider many signals, and the linked page still needs appropriate content and technical accessibility.
How Crawlers Encounter Internal Links
Search crawlers can follow crawlable HTML links from a source URL towards a destination URL. Consequently, links within navigation, breadcrumbs, page copy, and other accessible elements can provide routes to additional pages.
Discovery represents only one stage.
A crawler finding a URL does not guarantee that a search engine will index it. Likewise, indexation does not guarantee rankings for desired queries. Crawling, indexing, and ranking remain separate processes.
Technical conditions can also change what happens after discovery. A destination might redirect elsewhere, declare another canonical URL, return an error, restrict crawling, or carry indexing directives that affect its search treatment.
Therefore, an internal link should not serve as a substitute for checking the destination itself. Site owners should confirm that strategically important destinations return the intended response, remain accessible, and match their desired indexing configuration.
Internal Linking Shapes Website Architecture
A website’s architecture describes how its pages and sections relate to each other. Internal links help make those relationships navigable.
A relatively simple service website might connect its homepage with core service pages, sub-services, supporting resources, legitimate location pages, and contact pathways. An online shop may need categories, subcategories, products, filters, buying resources, and account functions.
No single hierarchy suits every website.
Nevertheless, important commercial and informational pages should normally have meaningful pathways from other relevant pages. Search crawlers and users should not need to rely solely on an XML sitemap to encounter important content.
Internal links can also connect conceptually related resources. A central resource covering a broad subject might reference narrower supporting pages where readers need additional detail. Supporting pages can link back where the broader resource provides useful context.
Such connections can reinforce logical topic relationships without requiring rigid cluster formulas.
Navigation and Contextual Links Serve Different Jobs
Primary menus provide predictable access to major areas of a website. Breadcrumbs show hierarchical position and give visitors a route towards broader sections. Footers may provide persistent access to important corporate, legal, support, or navigational destinations.
Contextual links operate differently. Writers place them inside page copy where another page can expand a specific point or provide the next useful step.
For example, an online shop’s navigation might link to a running-shoes category from every page. A shoe-care resource could separately link to that category when discussing suitable footwear products. The contextual link arises from the subject being discussed.
Neither placement carries a known fixed SEO value. Instead, each serves a different navigational and contextual function.
Anchor Text Gives Internal Links Meaning
Anchor text refers to the clickable wording associated with a link. Descriptive wording can help users anticipate the destination and provide search systems with context about the linked page.
A vague anchor such as “click here” reveals little by itself. A descriptive phrase naming the relevant service, product category, process, or resource usually communicates more useful information.
However, descriptive does not mean aggressively optimised.
Repeatedly forcing an identical commercial phrase into unrelated sentences can reduce readability and create unnatural copy. Writers should choose wording that accurately describes the destination while fitting the surrounding sentence.
Navigation naturally repeats anchors because menus use consistent labels. That repetition differs from repeatedly manufacturing exact-match contextual anchors throughout editorial copy.
Images can also function as links. Where an image acts as the clickable element, meaningful alternative text can help communicate the image’s purpose and provide contextual information when appropriate.
Internal Links and Link Equity
Search engines have historically used links as part of systems for assessing relationships and importance across the web. Within a website, internal links can help connect prominent pages with deeper relevant destinations.
However, website owners cannot reliably calculate exact ranking value passed through each internal link with a simple formula.
The source page, destination, broader architecture, relevance, crawlability, and other factors all affect the practical significance of a connection. Furthermore, adding dozens of unnecessary links does not create a predictable ranking gain.
Internal-link decisions should therefore start with information architecture and reader needs.
A prominent category page may reasonably connect to major subcategories. A highly relevant informational resource may provide contextual pathways towards related services. Conversely, forcing links from every high-visibility page towards every commercial URL can weaken clarity.
Link equity provides a useful conceptual model, but precise manipulation should not replace sensible architecture.
Orphan Pages and Deep Content
An orphan page lacks meaningful internal links pointing towards it. Such a URL may still appear in an XML sitemap, receive backlinks, or become known through other technical routes, but its isolation creates architectural problems.
Users browsing the website may have no natural route to reach it. Search crawlers may also receive fewer internal pathways towards the URL.
Not every isolated URL requires correction. Temporary campaign pages, utility URLs, or intentionally separated content can have different purposes. However, an important product, service, category, or evergreen resource usually deserves relevant internal connections.
Click depth raises a related issue. Strategically important pages should remain reasonably accessible, but no universal rule requires every URL to sit within a fixed number of clicks.
Large e-commerce sites, archives, marketplaces, and documentation platforms naturally develop deeper structures. The practical question concerns accessibility and importance, not compliance with an arbitrary number.
Integrating New and Existing Content
Publishing a new page should trigger internal-link decisions rather than end the content process. The new URL needs a sensible position within the existing architecture.
Editors can review older relevant pages and add contextual links where the destination genuinely expands the discussion. Category pages, resource hubs, related-content modules, or other established pathways may also warrant updates.
The new page should provide useful outbound internal pathways of its own. Linking back to relevant supporting resources can prevent it from becoming an isolated endpoint.
Older content also requires periodic maintenance. New resources may provide better destinations than links chosen several years earlier. Product availability can change, service structures can evolve, and consolidated pages can replace multiple outdated URLs.
After migrations or URL changes, teams should update internal links to point directly towards appropriate current destinations where practical. Redirects remain useful, but relying indefinitely on unnecessary intermediate hops complicates navigation and maintenance.
How External Links Work
An external link takes a visitor from the current website to another website. The linking page acts as the source, while the external URL becomes the destination.
Websites use these links for many legitimate editorial purposes. A technical resource might reference official specifications. A regulatory page might direct readers towards the responsible authority. Research-focused content might reference original evidence. A product resource could link to a compatible external tool where that destination helps users complete a task.
External linking therefore serves a different architectural purpose from internal linking.
The website owner controls the source link but does not control the external destination. Consequently, external links require periodic review because destination owners can remove pages, change content, introduce redirects, or allow domains to expire.
Context still matters. Anchor wording and surrounding copy should accurately represent what readers will encounter after leaving the site.
External Links Can Support Editorial Usefulness
External links can make content more useful when another source provides evidence or information that the website cannot reasonably reproduce.
A page discussing regulations, for example, may benefit from directing readers towards the relevant official material. Research claims can link to original research where appropriate. Technical specifications may warrant references to authoritative documentation.
Such links support verification and transparency. However, merely linking to respected websites does not automatically increase the source page’s rankings.
That distinction matters because external-link advice sometimes turns legitimate sourcing into a mechanical SEO tactic. Adding an unrelated link to a prominent domain simply because the domain appears authoritative provides little editorial value.
The external destination should earn its place through relevance.
Similarly, source linking alone does not create expertise, experience, authoritativeness, or trustworthiness. Editorial quality depends on the accuracy, usefulness, sourcing, authorship, and broader characteristics of the content and website.
Followed Links and Link Attributes
A normal link does not require a special rel value merely because it points externally. However, particular relationships can warrant link attributes that provide additional information about the nature of the link.
rel="sponsored" identifies links associated with advertisements, sponsorships, or other compensated arrangements. Current search documentation recommends this value for paid links.
rel="ugc" Can identify links within user-generated content, including comments and forum contributions.
rel="nofollow" Remains available where a publisher does not want to associate its site with the linked destination or prefers not to provide the normal form of endorsement. Search engines can treat such attributes as hints for ranking purposes rather than applying the older assumption that nofollow creates an absolute barrier to every form of processing.
Publishers can also combine appropriate values where circumstances require them.
The relationship behind the link should determine qualification. Site owners should not automatically apply nofollow to every external destination.
Paid Links Require Appropriate Treatment
Commercial relationships create different considerations from independent editorial references.
Advertising, sponsorship, compensated placements, and other paid linking arrangements should receive appropriate qualification. Current search policies allow paid links for legitimate advertising and sponsorship purposes when publishers qualify them appropriately, including through sponsored or nofollow.
The underlying distinction concerns ranking manipulation.
Buying or selling links specifically to manipulate search rankings can conflict with search spam policies when those links pass ranking credit. Therefore, businesses should not assume that a relevant destination makes an undisclosed paid link acceptable for ranking purposes.
Incorrect qualification does not justify predictions about automatic penalties in every case. Search systems and enforcement processes involve more nuance than such claims suggest.
Publishers should instead disclose commercial relationships appropriately and implement suitable technical attributes from the outset.
User-Generated Links Need Separate Consideration
Websites with forums, comments, reviews, community discussions, or other user-submitted material may contain links that editors did not independently select.
The ugc attribute provides a way to identify links within user-generated content where appropriate. Sites can also moderate submissions, restrict linking privileges, or apply other controls according to their community model.
This distinction matters because editorially chosen external references and links submitted by users represent different relationships.
Moderation remains necessary even when attributes exist. A forum filled with spam links still creates usability and quality problems. Technical qualification should support sensible publishing controls rather than replace them.
Internal and External Links Have Different Functions
The clearest comparison starts with destination and purpose.
Internal links keep users within the same website context. They support navigation, architecture, crawl paths, page relationships, and routes towards relevant products, services, resources, or actions.
External links take users elsewhere. They commonly support sourcing, evidence, attribution, specifications, independent resources, or useful destinations beyond the site’s own scope.
Website owners have greater control over internal destinations. They can change page copy, repair URLs, update navigation, modify anchors, and redesign architecture. External destinations remain under another publisher’s control.
Maintenance therefore differs. Internal-link audits focus heavily on architecture, orphan pages, redirects, hierarchy, and destination importance. External-link audits focus more on destination availability, continuing relevance, sourcing quality, redirects, safety, and commercial qualification.
Both types still require accurate anchor text and editorial relevance.
How Both Link Types Work Together
A well-developed page can use internal and external links without creating a conflict between them.
Consider a resource explaining product safety. An internal contextual link might direct readers towards a related product category or another page explaining maintenance. Meanwhile, an external link could point towards an official safety standard that supports a factual requirement.
The internal destination extends the user’s journey through the website. The external destination supplies information that appropriately belongs elsewhere.
Editors should evaluate each link independently. If an internal page provides the best next step, link internally. If an independent primary source provides necessary evidence, link externally.
This approach avoids treating links as quotas. Each connection earns its position through reader purpose and contextual relevance.
How Many Links Should a Page Contain?
No universal number determines the correct quantity of internal or external links.
A short commercial page may need relatively few contextual links because its purpose remains focused. A long technical resource may legitimately reference many supporting pages and external sources.
Information density also matters. If every sentence contains several links, readers may struggle to identify meaningful pathways. Conversely, a long resource that mentions numerous related concepts without providing useful internal connections may miss obvious navigation opportunities.
Navigation contributes additional links that exist independently of body copy.
External-link volume follows similar logic. Some pages need no external references because they make no claims requiring outside sourcing. Research-heavy or regulatory content may need several.
Editors should judge links by purpose, relevance, clarity, and usefulness instead of aiming for a predetermined count.
Placement Should Follow User Purpose
Links can appear in introductions, body paragraphs, lists, breadcrumbs, navigation, resource sections, calls to action, footers, and related-content modules.
Placement should match function.
A contextual link belongs close to the statement or concept it supports. Breadcrumb links belong within hierarchical navigation. Calls to action can direct visitors towards conversion pathways. Resource sections can group supporting destinations when that presentation improves scanning.
No reliable universal rule makes one location the strongest SEO position in every circumstance.
Prominence can affect how users encounter a link, but website owners should avoid converting placement into an artificial ranking formula.
A link near the beginning of a page can make sense if readers need that destination immediately. Otherwise, placing it there solely because of an assumed SEO advantage can harm editorial flow.
Broken Links Need Different Fixes
Broken internal links usually indicate a problem within the website owner’s control. Teams can correct the URL, restore missing content, redirect an obsolete destination towards an appropriate replacement, or remove the link if no useful destination exists.
External broken links require different decisions. Editors can locate an updated version of the source, replace it with another credible destination, or remove the reference when it no longer serves the content.
Redirects also require review.
An internal link that passes through several unnecessary redirects may still reach a functioning page, but linking directly to the intended final URL creates a cleaner path. Site migrations therefore provide a strong reason to update internal destinations after redirects have been implemented.
For external redirects, teams should verify the final destination. A once-reliable reference can eventually redirect towards unrelated or unsuitable content.
URL Consistency Simplifies Internal Linking
Internal links should generally point towards the preferred form of a destination URL where practical.
Inconsistent links involving HTTP and HTTPS, www and non-www hosts, trailing-slash variations, unnecessary parameters, or old redirected URLs can create avoidable complexity.
Canonicalisation remains a broader technical issue, and internal links alone do not determine which URL a search engine selects as canonical. However, consistently linking towards intended URLs supports clearer implementation.
Site owners should pay particular attention after migrations, redesigns, protocol changes, and CMS modifications.
Some internal destinations also legitimately remain non-indexable. Account areas, carts, login pages, administrative functions, and other utility pages can require internal links because users need them even when search indexing provides no value.
Therefore, an internally linked URL does not automatically need to become an indexable landing page.
E-Commerce Sites Need Scalable Internal Structures
Large online stores face internal-linking challenges that small informational websites rarely encounter.
Categories and subcategories create core browsing paths. Product pages connect users with individual items, while related-product modules can provide useful alternatives. Buying resources may connect informational demand with relevant commercial categories.
Faceted navigation adds complexity. Filters can generate many URL combinations, so stores need deliberate technical handling instead of allowing every possible filtered state to become an uncontrolled search landing page.
Inventory changes create further maintenance requirements. Discontinued products need appropriate handling based on replacement availability, ongoing demand, user value, and business requirements.
Internal linking should reflect useful shopping relationships. Linking every product to numerous unrelated categories merely to create more connections can make navigation less coherent.
External links remain appropriate when outside specifications, manufacturers, standards, or other resources genuinely support shoppers.
Service Websites Need Intent-Based Connections
Service businesses usually operate with fewer URLs, but relevance still determines useful internal pathways.
A core service page can connect to legitimate sub-services, supporting informational resources, appropriate case material, and enquiry pathways. Location pages can connect where geographical relationships genuinely help visitors.
Problems arise when businesses create large numbers of near-identical service-location pages and cross-link them mechanically. A visitor reading about one location rarely benefits from dozens of unrelated location anchors inserted for search purposes.
Informational resources should link towards services when the service represents a logical next step. Likewise, service pages can reference educational material when visitors need additional detail before making a decision.
Content clusters can support this model without becoming rigid templates. Central resources and supporting pages should connect where reader intent overlaps, not simply because every page belongs to a predefined topic group.
Common Internal Linking Problems
Internal-link problems often emerge gradually as websites publish, migrate, remove, and reorganise content.
- Orphan pages: Important URLs lack internal pathways, weakening navigation and making architecture less coherent.
- Broken links: Users and crawlers reach missing destinations instead of the intended resource.
- Generic anchors: Repeated vague wording provides limited destination context.
- Redirected destinations: Old URLs create unnecessary intermediate steps when direct links could point towards current pages.
- Irrelevant cross-linking: Pages connect despite weak topical or navigational relationships.
- Link-heavy paragraphs: Excessive links make copy difficult to scan and obscure useful pathways.
- Weak hierarchy: Important categories or services sit behind unclear navigational routes.
- Low-value destinations: Repeated links direct attention towards pages that provide little reader benefit.
- Migration leftovers: Internal links continue using obsolete URL structures after website changes.
Regular review helps identify these issues before they spread across templates or large content libraries.
Common External Linking Problems
Outbound-link problems usually concern destination quality, continuing relevance, or the relationship behind the link.
- Weak sources: A page references secondary or unreliable material when stronger primary evidence exists.
- Outdated destinations: The linked information no longer reflects the claim being supported.
- Undisclosed commercial relationships: Paid or sponsored links lack appropriate treatment.
- Irrelevant references: External links appear without helping readers verify or expand a point.
- Excessive affiliate linking: Commercial destinations overwhelm the editorial purpose of the page.
- Broken references: Linked resources disappear without editors updating the source page.
- Misleading anchors: Clickable wording creates inaccurate expectations about the destination.
- Unreviewed legacy links: Old external destinations change ownership, redirect elsewhere, or deteriorate in quality.
External-link maintenance therefore belongs within normal editorial maintenance, especially for evergreen resources that remain published for long periods.
Linking Myths That Should Not Become Rules
Several popular recommendations turn conditional practices into fixed formulas.
A page does not need a predetermined number of internal links. Likewise, every informational page does not require an external reference.
External links do not automatically reduce rankings simply because visitors can leave the website. Conversely, linking to a prominent or respected domain does not automatically improve rankings.
Applying nofollow to every outbound link also misunderstands link attributes. Publishers should select attributes according to the relationship involved.
Exact-match internal anchors do not guarantee better performance. Descriptive wording should remain accurate and natural.
Similarly, no universal click-depth number suits every website architecture. Important pages should remain reasonably accessible, but complex sites naturally require different structures.
Finally, more links do not automatically create stronger SEO. Every additional link should serve a meaningful navigational, contextual, sourcing, or functional purpose.
Auditing Internal and External Links
A useful audit separates architectural problems from outbound-reference problems. Combining everything into one link count can hide the issues that actually require action.
Internal Audit Checks
Teams should review:
- important pages with few or no internal links;
- orphan URLs found through sitemaps or other inventories;
- broken internal destinations;
- links passing through unnecessary redirects;
- anchors that poorly describe destinations;
- obsolete navigation;
- inconsistent preferred URL versions;
- irrelevant cross-links;
- pages overloaded with low-value links;
- older content lacking links towards useful newer resources.
Prioritisation matters. A broken link within a prominent navigation template can affect many pages, while an isolated outdated link may require less urgent attention.
External Audit Checks
Outbound reviews should verify that destinations still load, remain relevant, and continue supporting the associated statement.
Editors should inspect unexpected redirects, changed domains, obsolete research, questionable resources, inaccurate anchors, and commercial relationships requiring appropriate qualification.
External-link auditing differs from backlink auditing. The former reviews links leaving the website. The latter evaluates links arriving from other websites.
That distinction should remain explicit in reporting so teams do not confuse editorial maintenance with off-page backlink analysis.
Choosing the Right Destination
Internal-link selection should start with the reader’s next likely need.
A useful destination might provide deeper explanation, a related service, a relevant category, a complementary product, supporting evidence, or a necessary action. The source and destination should share a clear contextual relationship.
Writers should not insert links solely because a particular page appears to need additional link equity.
External links require a parallel editorial test. Another website may provide original evidence, official requirements, technical specifications, primary research, or specialist information beyond the source site’s reasonable scope.
Sometimes no outbound link adds meaningful value. If a page makes a straightforward statement that does not require external verification, adding a random reference simply to satisfy an SEO checklist creates clutter.
The destination should justify the click.
Evaluating Professional SEO Support
Businesses seeking professional assistance should evaluate how a provider approaches architecture and linking rather than focusing on promises about rankings.
Someone searching for the best seo service provider should examine whether prospective specialists can audit crawl paths, identify orphan pages, analyse internal-link relationships, diagnose redirect problems, assess content relevance, and explain recommendations transparently.
Useful support should connect technical findings with business priorities. For example, an audit should distinguish strategically important pages from low-value utility URLs instead of recommending identical treatment for every page.
Reporting should also show what changed, why teams made the change, and what indicators they will monitor afterwards.
Realistic claims matter because no provider can guarantee a specific ranking improvement from adding a particular set of internal or external links.
A Practical Linking Framework
Before adding, retaining, or changing a link, editors and SEO teams can apply a short decision process:
- Does the destination genuinely help the reader at this point?
- Does the anchor accurately describe what follows?
- Does the link fit the surrounding sentence and page intent?
- Does the destination URL work correctly?
- Would an existing internal resource satisfy the need better?
- Does an external source provide evidence or information the website should not duplicate?
- Does payment, sponsorship, or user-generated content require suitable qualification?
- Can users reasonably reach strategically important internal pages?
- Do outdated links require replacement or removal?
- Does the connection strengthen useful architecture instead of creating clutter?
These questions keep linking decisions tied to purpose. Technical SEO then supports those decisions by ensuring that the resulting pathways function as intended.
Conclusion
Internal links organise and connect content within a website, while external links connect readers with relevant resources beyond it. Their SEO functions differ, yet both depend on contextual relevance, accurate destinations, useful anchor text, sound technical implementation, and maintenance. Internal linking should create sensible architecture and navigational pathways. External linking should support sourcing, verification, or useful onward journeys where appropriate. Instead of applying fixed link counts or mechanical formulas, evaluate every connection by one practical standard: the link should have a clear reason to exist for the user.
FAQs
What is the main difference between internal and external links?
Internal links connect pages within the same website context, helping users and crawlers move through its content. External links point from that website towards another website. A backlink differs because it points from another website towards the site being evaluated. Each relationship serves a different SEO and editorial function.
Do internal links help SEO?
Internal links can support SEO by creating crawl paths, connecting related pages, clarifying website architecture, and providing destination context through anchor text. However, adding links does not guarantee ranking improvements. The destination still needs relevant content, appropriate technical configuration, and a meaningful role within the wider website structure.
Can external links harm SEO?
Relevant editorial links to useful external resources do not inherently create an SEO problem. Problems can arise from manipulative linking, unsuitable commercial arrangements, compromised destinations, misleading anchors, or low-quality references. Publishers should choose destinations carefully, qualify commercial links appropriately, and periodically verify that long-standing external references remain suitable.
How many internal links should a page contain?
No universal number applies. Appropriate link volume depends on page length, information density, navigation, website architecture, and the number of genuinely useful destinations. A page should contain enough internal links to support logical navigation and related content without overwhelming readers with unnecessary or repetitive clickable elements.
Does every page need an external link?
No. External links should serve an editorial or functional purpose rather than satisfy a fixed SEO requirement. A page may need external sources for research, regulations, technical specifications, or verification. Another page may communicate its purpose fully without sending readers elsewhere. Context should determine the decision.
What anchor text works well for internal links?
Descriptive, concise wording that accurately represents the destination usually serves users effectively. The anchor should fit naturally within the surrounding sentence. Writers should avoid repeatedly forcing identical commercial wording into unrelated contexts. Navigation anchors can repeat naturally because menus use stable labels for consistent destinations across the website.
What is an orphan page in SEO?
An orphan page has no meaningful internal links pointing towards it. Search engines may still encounter the URL through an XML sitemap, backlinks, or other routes, but users can struggle to reach it through normal navigation. Important orphan pages usually warrant relevant internal connections unless intentional isolation serves a specific purpose.
Should all external links use nofollow?
No. Publishers should choose link attributes according to the relationship involved. Standard editorial links do not require it nofollow merely because they lead elsewhere. Sponsored or compensated links require appropriate qualification, while ugc can identify links in user-generated material. nofollow remains available for suitable circumstances but should not become an automatic default.
Do broken links affect website quality?
Broken links create poor navigation because users cannot reach the intended destination. Broken internal links can also disrupt useful crawl paths and reveal outdated architecture. External broken links weaken sourcing and usability. Website owners should repair, replace, redirect, or remove broken destinations according to the purpose and ownership of each link.
Can a page contain too many internal links?
A high link count is not automatically problematic. However, excessive or irrelevant links can make content difficult to scan, dilute navigational clarity, and send users towards destinations they do not need. Editors should assess each link’s purpose and remove repetitive connections that contribute little to navigation, context, or task completion.