Small sites rarely have an internal linking problem in the sense of quantity. Most have enough links. What they lack is placement: the links exist in a footer column or a sidebar, in the part of the page that nobody reads and that search engines treat as generic.
Internal linking is the cheapest form of SEO because it requires no new content and no outreach. It also has an unusually clear mechanism: a link passes context and attention from one page to another, and where that link appears determines how much of both arrive.
Why the link count is the wrong metric
Adding links to a site-wide footer template is not internal linking. It is a global navigation element, and each link carries almost no contextual signal because the same link appears on every page regardless of relevance.
A link inside the body of an article is different in kind. It sits in a paragraph that establishes context, it appears on a page whose topic is related, and it is clicked by readers who have just demonstrated interest in the subject. That combination is what makes it valuable.
A useful working rule: one contextual body link is worth more than a hundred template links. This is why the useful unit of work is not "add more internal links" but "find the paragraphs that should contain one and do not".
Link from the body
The placement question has a clear answer in most cases: inside the main content, in a paragraph where the target is genuinely relevant to what the paragraph is discussing.
Two placements to avoid. The first is a block of related links at the bottom of the page — it works for readers who finish, but it carries less context than an in-line mention and it is easy to ignore. The second is navigation, which is for orientation rather than for linking concepts.
There is a second reason body links matter more. Search engines use the words around a link as part of the signal. A link sitting inside a sentence inherits the meaning of that sentence. A link in a list of five related titles inherits almost nothing.
Anchor text that describes the destination
Anchor text should describe what the destination page is about, in words a reader would find useful. "Read more" and "click here" tell neither the reader nor the algorithm anything.
Two failure modes to avoid in the other direction. Using the exact same anchor text every time you link to a page looks manipulative and produces an unnaturally repetitive reading experience. Using an anchor so long that it is a full sentence dilutes the signal across too many words.
The natural version is what you would say if you were describing the page to a colleague: two to six words, specific to the destination, varied across pages.
The hub-and-spoke pattern
The most useful structure for a small site is a hub page that covers a subject broadly, with spoke pages that go deep on each part of it. Links run in both directions: the hub links to each spoke from the paragraph that introduces that part, and each spoke links back to the hub, usually in the first or last section.
The benefit is directional. The hub concentrates whatever authority the site has and distributes it to the spokes; the spokes return the reader to the hub rather than letting them leave. Without the hub, a site is a set of disconnected pages that each have to earn everything they get.
A small site needs only two or three hubs to see the effect. Trying to build one for every category before any of them has content is how the structure ends up unused.
Linking new pages backwards
When you publish a new page, the most valuable thing you can do in the first hour is find the three or four existing pages that should link to it, and add those links.
Two reasons. A new page with no internal links pointing at it is an orphan, and it will be discovered slowly if at all. And the older pages usually have more accumulated authority to pass; being linked from an established page is worth considerably more than being linked from a page that was published the same week.
It is worth keeping a simple habit: before publishing, list the existing pages that ought to reference the new one, and edit them in the same session. Doing it later rarely happens.
Finding orphans
An orphan page has no internal links pointing to it. It exists in the sitemap and possibly in search results, but not in the site's own structure.
Orphans accumulate quietly. They appear when a page is published and never referenced, when a template changes and drops a link, or when a page is renamed and old links are not updated.
A simple way to find them: for each URL in the sitemap, count how many other pages link to it. Anything with a count of zero, or only a link from the homepage or a footer, needs attention. These tend to be exactly the pages that perform worst, and often the fix costs one paragraph elsewhere.
How many links per page
There is no correct number. The practical constraint is whether each link is there because it helps the reader at that point, or because the page was felt to need more links.
Two or three contextual links in a thousand-word article is a reasonable expectation. A page with twenty in-body links is usually a list rather than an article, and the links are competing rather than reinforcing each other.
The same logic applies to the whole site. Adding links everywhere dilutes each one. Concentrating them on the pages that matter is more effective than distributing them evenly, and it is easier to maintain — a smaller number of deliberate links can be checked and updated.
An audit you can run in an hour
For a site under a hundred pages, the whole audit fits in an hour with nothing but a spreadsheet and the site open.
- List every page URL.
- For each one, note how many internal links point to it, and from where.
- Mark every page with zero links, or with links only from the homepage and footer.
- Pick the three pages you most want to rank, and check whether anything links to them from body content.
- For each weak page, note the two existing pages where a natural mention belongs.
The output is a short list of edits, each of which is one sentence added to a page that already exists. That is the entire deliverable, and it is worth more than most of what gets published for the sake of having a new page.