Experiment
Internal Linking Density vs. Ranking Velocity
Increasing internal link density to a target page by 3x (from ~5 internal links to ~15) will produce measurable ranking improvements within 60 days, independent of new backlink acquisition.
What is being tracked.
How the test is structured.
We selected 10 pages across 2 test domains in the health vertical, all ranking in positions 11 to 20 for their primary keyword. For 5 pages (test group), we added 10 additional contextual internal links from topically relevant pages. The other 5 pages (control group) received no changes. No external link building was performed during the test period.
What has happened so far.
The original entry reported a 4.2-position average improvement for five test pages, compared with 0.8 positions for five control pages. It also reported that three test pages reached page one and crawl frequency rose 40 percent.
September 11, 2026 evidence audit: I cannot find the page list, query list, raw ranking series, crawler extract, variance, or statistical test behind those numbers in the surviving site or warehouse records. The result may have been recorded accurately at the time. I cannot reproduce it now. With five pages in each group, calling it statistically significant without the actual test was stronger than the evidence I can still show.
What would support or challenge it.
Status: unverified historical result.
The reported numbers remain here as a dated receipt, but the claim of statistical significance is withdrawn. The defensible conclusion is smaller: the observation was consistent with stronger internal linking helping pages already near page one, but this record cannot prove the effect or separate it from other changes.
If I run this again, the page set, queries, change log, daily positions, crawler counts, and test method get stored before the conclusion gets written.
Keep Following the Tests
Move from this open thread back into the full experiment library.