Internal linking gets filed under technical housekeeping. For B2B SaaS, it is an acquisition decision. A brand can have useful content and still make it unnecessarily difficult for people—and the systems helping them—to find the next relevant page.
Key figures in text
| Measure | Value |
|---|---|
| Reported organic sessions over the modeled comparison | +32.3% |
| Measurement window | March 27–April 23 (four weeks) |
Start with the structure already on the site
The default response to a content gap is often another brief. My first question is different: do we need another page, or do we need a better connection to something that already exists?
A useful product explanation, comparison or implementation guide can sit several steps away from the page where a buyer needs it. That is a journey problem as much as a search problem. Internal links should make the relationship between those pages explicit.
A link graph, not a handful of recommendations
In the work described in my original analysis, we used the Quattr API to rebuild internal connections across a B2B site. The approach connected related content through browse modules and semantic phrases informed by non-brand search demand.
The important distinction is between recommending links and maintaining a useful linking system. The latter needs a defined page population, sensible destinations, appropriate anchor language and a repeatable implementation path. AI can help propose those connections; the source data and publishing rules keep the work grounded.
Make the measurement part of the implementation
The chart follows organic sessions before and after the linking change, alongside the study’s modeled comparison. It reports 32.3% growth over expected sessions in the four-week measurement window.
That is a more useful conversation than whether the team added a large number of links. It brings the discussion back to the pages affected, the period observed and the outcome the work was meant to influence.
Where I would start
Choose one commercially useful page family. Work outward from the questions a buyer is trying to resolve, then examine whether the site makes the relevant answers easy to reach.
- Map existing pages to demand and the next useful buyer action.
- Identify missing connections and test whether each destination genuinely helps.
- Implement a bounded set of changes with named owners and acceptance checks.
- Review search, referral and conversion measures separately, against their own baselines.
Connect what you already have before you write more.
Source & measurement notes
The original graphic compares sessions on linked pages with a modeled no-change baseline. Its reported result is +32.3% over four weeks. This figure covers organic sessions; the other outcomes discussed in the original post are separate measures.
Read my original analysis ↗