Internal Linking Services That Move Equity to Pages That Sell
Your strongest pages are already passing authority somewhere. The question is whether it lands on a page that converts or on a tag archive nobody reads. We draw the real route, cut the dead ends, and connect every money page to the hub that feeds it.
One crawl, one map, one prioritised list of links to add — before any retainer talk.
Your best page is probably feeding a page that earns nothing
Internal links are the only ranking input you control completely, cost nothing per unit, and can be changed this afternoon. Most sites still treat them as decoration: a "related products" widget that points at whatever was published last, a footer with forty links to policies, and a navigation menu that has not changed since launch.
The result is a network where authority pools in the wrong places. The homepage sends equity to five category pages, those pages link to every product equally, and the one commercial page that actually closes deals sits four clicks deep with two inbound links — both from a blog post published in 2021.
Three symptoms show up again and again:
- Pages that get impressions but almost no clicks because they are buried below the third click depth
- Clusters that exist in a sitemap but not in the link graph — articles that never reference the service page they support
- Anchor text that is either identical across two hundred links or meaningless, like "click here" or "read more"
Internal linking services are not a link-building substitute. They are the plumbing that decides whether the authority you already earned reaches the page that needs it.
What our internal linking services deliver
A crawl, a map, a policy and a ticket list — four artefacts your developer can act on without a workshop.
Route map
Every indexable URL placed by click depth, with inbound and outbound links counted. Hubs, dead ends and one-way pages are marked in the diagram.
Cluster plan
Which page is the hub for each topic, which pages support it, and the exact internal links that connect them in both directions.
Anchor policy
A short rule set for each destination: primary phrase, acceptable variations, and the phrases nobody should use because they misdescribe the page.
Link tickets
Source URL, target URL, anchor, paragraph location and priority. Written for a CMS editor, not for a strategist.
Orphan recovery
A list of pages with zero inbound internal links, plus the three most sensible places to link them from.
Measurement sheet
Baseline and follow-up readings for click depth, indexed pages, impressions per cluster and organic sessions on the target set.
Equity follows routes, not good intentions
A link is a vote with a direction. Pages with more internal votes, closer to the homepage, receive more of the crawl and more of the authority signal. That is why the single most effective change on most sites is not adding links everywhere — it is redirecting the links you already have.
We start from the homepage and follow the graph outward, recording depth for every URL. Anything past depth three is treated as a candidate for promotion. Then we look at where the commercial pages sit and how many contextual links point at them, ignoring navigation and footer, because those two blocks appear on every page and are discounted by everyone including the crawler.
From there the work is arithmetic. If a product page needs to rank and has two inbound contextual links, we plan for eight to twelve from relevant pages. If a policy page has ninety inbound links because it sits in the footer, we do not touch it — we simply stop counting it as equity.
Clusters and hub pages
A cluster is a hub page plus the pages that support it. The hub targets the commercial phrase and links out to every supporting page; each supporting page links back to the hub and to two or three siblings. That shape gives the crawler a clear path and gives the hub a concentration of internal votes that a flat blog archive never accumulates.
The most common structural mistake is a hub that exists in the menu but is never referenced in body copy. The second is a cluster where every article links to the hub but the hub links to none of them, so the supporting pages receive nothing.
Anchor text has to describe the destination
Google reads the words around a link to understand what the target page is about. That makes anchor text a description, not a ranking trick. A page about stainless steel flanges should be linked with phrases that describe stainless steel flanges, varied naturally as a writer would vary them.
The failure modes are symmetrical. Two hundred identical exact-match anchors look engineered and read badly. Two hundred "learn more" anchors tell the crawler nothing and waste the strongest signal on the page. The middle path is a small set of descriptive variants used where they fit the sentence.
| Source page | Destination | Anchor used | Why this one |
|---|---|---|---|
| Hub: valve repair | Service page | valve repair and refurbishment | Primary phrase, placed in the first screen of body copy |
| Case study | Service page | on-site gate valve rebuilds | Specific variant that matches the case facts |
| Guide: selection | Category page | stainless steel valve range | Describes the catalogue, not the brand |
| Comparison | Product page | the 3-inch flanged model | Narrow, useful to a buyer mid-decision |
We write the policy before the tickets, because otherwise an editor will fill a table with the same phrase forty times and call it optimisation. Getting the anchor set right is part of the on-page layer, which is why it runs alongside On-Page SEO Services on most engagements.
Orphan pages: indexed, unreachable, invisible
An orphan page is a URL that exists, is often in the sitemap, sometimes ranks a little — and has no internal links pointing at it. The crawler can reach it through the sitemap, so it may be indexed. Buyers cannot reach it, so it earns nothing. It is the most wasteful state a page can be in.
Orphan diagnostic · /guides/pressure-rating
- Inbound internal links
- 0
- Click depth from home
- 5
- Present in sitemap
- yes
- Found by crawl
- yes
- Impressions, 90 days
- 1,240
- Sessions, 90 days
- 11
Verdict
A page that earns impressions and no visits is a page with demand and no route. The fix is not a rewrite — it is three contextual links from pages that already discuss pressure ratings, one link from the hub, and a place in the navigation for the series. We log the change, then compare click depth and sessions on the same URL after four weeks.
Orphans usually group by cause: legacy URLs from a migration, pages created by a product feed, and articles published without a cluster to join.
Too many links is also a routing problem
The old "no more than one hundred links per page" rule was always a proxy for a real problem: when a page links to everything, it endorses nothing. A category page with 300 product links, a footer with 80 policy links and a "related" module with 18 items dilutes every vote to the point where none of them mean much.
Our default is a budget per template. Navigation stays lean, the footer holds the legal minimum, and body copy carries the links that matter. On a category page, we usually keep the product grid plus three to five editorial links into the buying guides. On an article, two to five contextual links is enough to be useful without turning the paragraph into a directory.
The opposite failure is just as common: a 3,000-word guide with zero internal links, because nobody owned the job of adding them. Both are routing problems, and both are fixed with the same map.
Navigation and breadcrumbs still count
Breadcrumbs are the one place where a repetitive link is correct, because they describe hierarchy rather than topic. We keep them crawlable, marked up with structured data, and consistent with the URL path. The main navigation we treat as a fixed budget: if a new item goes in, something comes out, or the equity spread changes without anyone noticing.
How the rework runs, stop by stop
-
Crawl and draw the current line
Full crawl with click depth, inbound and outbound counts, plus the internal link graph exported so nothing is based on a screenshot. This is the baseline every later claim is measured against.
-
Choose the hubs
For each commercial theme, we pick the page that should own the phrase. Sometimes that is an existing category; sometimes it is a page that needs to be created and linked before it can rank.
-
Write the anchor policy
Primary phrase, allowed variants and banned phrases per destination, so editors can add links later without breaking the pattern or stuffing one phrase.
-
Ship link tickets in batches
Twenty to forty links per release, each with source, target, anchor and the paragraph it belongs in. Batches keep the change reviewable and the impact attributable.
-
Measure depth and demand
Four weeks later: click depth of the target set, indexed pages, impressions per cluster and sessions on the orphan list. Then the next batch.
What the map changed
"An industrial pump supplier, 6 months: 41 orphan pages connected, average click depth 4.2 to 2.1, organic inquiries 5 to 23 per month."
Pumps · US"A packaging converter, 4 months: one hub rebuilt with 22 supporting pages, category impressions up 64% and demo requests 8 to 21 per month."
Packaging · UK"A tools exporter, 3 months: footer cut from 74 to 19 links, product page equity recovered, organic revenue per session up 26%."
Hand tools · DEScope is based on the map, not on a page count
A small catalogue with a clean hierarchy needs a crawl, a policy and one batch of tickets. A site with 40,000 URLs, faceted navigation and a blog that has been publishing weekly for six years needs a graph analysis, a pruning plan and a rolling retainer.
We quote on the size of the link graph, the number of templates involved and whether content has to be written to create natural link placements. No per-link pricing, no links sold as a commodity — this is structure work, and it is priced like structure work.
If the crawl also turns up redirect chains, duplicate canonicals or index bloat, we flag them in the same report. Fixing those is often the cheapest way to help the links you are about to add, which is why the work pairs naturally with an SEO Services review.
Send us your sitemap
We will crawl it and return the three pages with the worst inbound-link situation, plus the single link that would help most.
Questions from the platform
How long before internal links show results?
Recrawling of the affected pages usually happens within two to five weeks. Depth and indexation changes appear first, impressions and rankings follow, and commercial metrics take a full quarter on competitive phrases.
Is this the same as building backlinks?
No. This is entirely on your own domain. Nothing is bought, placed or outsourced, so there is no link-scheme risk and no recurring cost per link.
Will adding links change how the pages read?
They should improve it. Every ticket specifies a sentence the link belongs in, so the result reads like an editor wrote it. If a placement feels forced, we drop it rather than stuff it into a caption.
What do you need from our team?
A crawl allowance, read-only Search Console access and one editor who can publish. If we have CMS access we place the links ourselves and log each change.
Can you work with a headless or JavaScript front end?
Yes, provided links are real anchors in the delivered HTML. If navigation is rendered client-side and crawlers only see a shell, that is a rendering issue we fix first — otherwise every link we add is invisible.
How do you measure success?
Click depth of the target set, count of orphan pages, impressions per cluster, and organic sessions and inquiries on the pages we linked to. We report the baseline, the change and the date of each batch.
Do we keep the map afterwards?
Yes. The crawl export, the cluster plan and the anchor policy are yours. We also include a short guide so a new editor can add links without breaking the pattern.
Who draws the line
A small team that has rebuilt link graphs for industrial suppliers, packaging converters and tool exporters — sites where the catalogue is large, the hierarchy is inherited from an old migration, and nobody has looked at click depth since launch. We work from crawl exports rather than impressions, and we write tickets an editor can finish in an afternoon.
Every recommendation comes with the number it should move. If we cannot name the metric, the link does not go on the list.