We Analyzed 312 Guest Post Placements. Here's What Actually Moved Rankings

Every link vendor claims their placements move rankings. We wanted to stop claiming and start measuring, so we tagged 312 guest post placements delivered across six client accounts between May 2024 and January 2026, and tracked the target pages they pointed at for 90 days each. Same reporting cadence, same rank tracker, no cherry-picking.

The short version: 61% of placements produced measurable lift for the target page. The other 39% sat there like decorative furniture — live, indexed, and completely inert. The interesting part is which 61%, because it wasn't who the metrics said it would be.

First: referral traffic on the linking page. Not the site's traffic — the page's. Pages sending even 5–10 visitors a month correlated with lift almost everywhere we looked. Pages on 'high traffic' sites that sent nobody correlated with almost nothing. Google isn't stupid: a page nobody visits through Google is a page Google doesn't trust with a recommendation.

Second: paragraph-level relevance. When the linking paragraph discussed the same topic as the target page — not just the same broad niche — lift was roughly double the average. Domain-level topical match without paragraph match barely outperformed the 39% dead zone. This is why our contextual screening now starts below the fold of the article, not at the domain.

Stop shopping by DR floors and start shopping by page behavior. Ask any vendor two questions: 'Will the linking page rank for something my buyers search?' and 'Will the paragraph actually be about my topic?' If the answer is a shrug, the DR number is doing all the selling, and the number doesn't rank you — the page does.

The boring conclusion nobody wants to pay for: five placements on pages that rank and send traffic beat twenty placements on impressive dashboards. We now price and plan campaigns around that ratio, and the 2025 cohort is outperforming the 2024 one by a wide margin.