We have spent £22,000 on link building across two agencies in 18 months. Our target keywords have barely moved. We cannot tell if link building is not working or if our agencies are not working.” That question is what led to an unusual arrangement: one client agreed to run four link building agency UK providers in parallel, on the same domain, for 6 months, then compare the outcomes against the metrics each agency reported.
The headline finding was not what any of the four agencies would have predicted. Domain Rating, the number every agency leads its pitch with, barely separated the winners from the losers. Organic traffic on the actual linking page did.
Here is what the test measured, what the data actually shows and does not show, and the checklist to run before spending another pound on links.
Key Takeaways
- Four agencies ran in parallel on one domain for 6 months, placing 116 links combined. The agency with the highest average linking-page traffic (1,200 sessions/month) drove the largest ranking movement; the agency with the lowest traffic (90 sessions/month) drove a negative movement, despite a respectable DR 38 average.
- Ahrefs’ own 2025 study found Domain Rating has roughly a 0.131 correlation with SERP position, a weak positive relationship, and Ahrefs’ own help centre states there is no evidence any search engine uses DR or a similar score as a ranking factor.
- Across the 116 links, pages under 200 organic sessions per month showed no statistically significant ranking correlation, while pages with 500 or more showed a consistent positive one, according to RankFast’s own analysis of the test.
- Editorial links, placed in newly published content, outperformed niche edits, inserted into existing content, by 34 percent on ranking impact per link in this test, despite similar DR and traffic profiles.
What Happened When Four Agencies Ran Link Building In Parallel For Six Months
Four agencies, anonymised as A through D, each worked the same client’s domain simultaneously for 6 months with no knowledge of each other’s placements, campaigns, or reporting. Combined, they placed 116 links. Each agency’s pricing, average Domain Rating, and reported ranking impact were tracked independently and compared at the end of the period.
The client kept the arrangement itself confidential from all four agencies for the duration of the test, which matters methodologically: an agency that knows it is being benchmarked against three competitors on the same domain has an incentive to perform differently than it would on a normal engagement. None of the four agencies were aware another agency was working the same domain until after the 6-month window closed.
| Agency | Price Per Link | Avg. DR | Links Placed | Avg. Linking-Page Traffic | Ranking Movement |
| A | £580 | 58 | 12 | 1,200 sessions/mo | +2.1 positions |
| B | £190 | 44 | 31 | 340 sessions/mo | +0.4 positions |
| C | £280 | 51 | 22 | 890 sessions/mo | +1.6 positions |
| D | £95 | 38 | 51 | 90 sessions/mo | -0.2 positions |
Read carefully before drawing a conclusion from this table alone: price, DR, traffic, and ranking movement are all in the identical rank order across these four agencies (A, then C, then B, then D, for every column). That means this summary, by itself, cannot separate which variable actually caused the ranking difference. Agency A could have won because of its traffic, its DR, its price, or some combination the four-row average cannot isolate. The next section covers what the underlying 116-link data is reported to show once traffic and DR are examined link by link rather than agency by agency, which is the only way to actually test the claim this piece makes.
The Metric That Actually Predicted Ranking Movement
According to RankFast’s analysis of all 116 links placed across the four agencies, link building results correlated with the organic traffic on the linking page itself, not with that page’s Domain Rating. Links from pages with under 200 organic sessions a month showed no statistically significant ranking correlation. Links from pages with 500 or more sessions showed a consistent positive correlation. Links from pages with 2,000 or more sessions showed the strongest per-link impact, and this pattern held across all four agencies regardless of their average DR.
| Linking-Page Traffic | Ranking Correlation Observed |
| Under 200 sessions/month | No statistically significant correlation |
| 500+ sessions/month | Consistent positive correlation |
| 2,000+ sessions/month | Strongest per-link impact observed |
These thresholds are RankFast’s own reported analysis of the 116-link dataset. I have not independently verified them, since the underlying link-level export was not available to check.
The practical implication, if the underlying data holds up: a DR 40 page with 1,500 monthly visitors is a better link than a DR 65 page with 80 visitors, for ranking purposes specifically. That is the opposite of how most agencies price a placement.
Why Domain Rating Is What Agencies Sell, Not What Predicts Rankings
Domain Rating is Ahrefs’ own metric, built entirely from the size and quality of a site’s backlink profile, and Ahrefs is direct about what it does not measure: content quality, topical relevance, or traffic. In Ahrefs’ own 2025 study, DR showed roughly a 0.131 correlation with SERP position, a real but weak relationship (Ahrefs). Ahrefs’ own help centre goes further, stating there is no evidence that Google or any search engine uses Domain Rating, or a similar third-party authority score, as a ranking factor, and cites Google’s John Mueller confirming Google does not have anything resembling a website authority score (Ahrefs Help Centre).
None of this makes DR useless. It is a genuinely useful shortcut for gauging backlink strength and prioritising prospects. It is a poor stand-in for what actually determines a link’s value to rankings, which is closer to whether real people find and click the page the link sits on. A link building agency UK businesses can evaluate leads with DR because it is easy to report on a single number in a monthly deck. It is harder, and more honest, to report on linking-page traffic, since that number is less flattering for agencies placing links on high-DR pages nobody actually visits.
There is a structural reason DR persists as the industry’s default metric even though its own creator downplays it: DR is visible before a link is placed, since it can be pulled from any domain in seconds, while traffic on the specific linking page often is not disclosed until after the campaign runs, if it is disclosed at all. A metric that is easy to check before signing a contract will always be easier to sell against than one that only becomes visible after the money has changed hands.
Editorial Links Vs Niche Edits: What The Test Showed
Editorial link building places a link within newly published content on an active publisher. A niche edit inserts a link into content that already exists and was published earlier. In this test, editorial links from comparable publishers outperformed niche edits by 34 percent on ranking impact per link, despite similar DR and traffic metrics between the two placement types.
The working theory is a freshness signal: new content on an active publisher tends to get crawled faster than an edit made to a page that has not been updated in months, and niche edits into genuinely stale pages, no updates in 6 or more months, showed the weakest performance of any placement type in the test. Niche edits are not obsolete as a tactic; a niche edit into a page that itself gets updated regularly performs closer to an editorial placement than to a stale one.
The practical filter for evaluating a proposed niche edit is the same traffic check used elsewhere in this piece: pull the linking page’s own traffic history, not just the domain’s overall DR, and check when that specific page was last meaningfully updated, not just when the site as a whole last published something new. A page that gets edited regularly and holds steady or growing traffic behaves, for ranking purposes, much closer to a fresh editorial placement than to the stale, forgotten pages that produced the weakest results in this test.
How To Audit A Link Building Agency Before Hiring Them
The question of how to evaluate link building agency proposals honestly starts with requesting a sample campaign report from a recent client in a comparable industry, then checking the underlying data yourself rather than trusting the summary the agency chose to send.
Most agency reports are built to flatter the agency, not to expose weak placements. That is not necessarily dishonest; it is simply what happens when the reporting template was designed around the metric that is easiest to make look good. Pulling the raw referring-domain list yourself, rather than reading the summary page, is the only way to see what the report was built to avoid showing.
| What To Check | Why It Matters |
| Organic traffic trend on each referring domain | A site losing traffic since 2023 is a red flag regardless of its DR |
| Topical relevance of the linking page to your niche | An unrelated but high-DR page passes less real signal than a relevant, lower-DR one |
| Anchor text diversity | More than 30 percent exact-match anchors is a penalty risk, not a quality signal |
| Whether the same page links to 8 or more other clients in similar niches | A clear paid link placement pattern that risks devaluation or penalty |
An agency that provides link reports showing only Domain Rating, with no traffic column, is not necessarily hiding something deliberately, but the omission is exactly the one that makes weak link quality hardest to spot.
What White Hat Link Building Actually Looks Like In 2026
Legitimate white hat link building in 2026 means earning or placing links on pages that carry a real audience, not just a favourable metric. That includes editorial placements on active publishers, well-executed niche edits on pages that themselves stay current, and digital PR that earns coverage on its own merits. Link building services 2026 pitches that lead exclusively with DR guarantees, or that cannot show a traffic column when asked, are optimising for the number that is easiest to report, not the one this test found actually moved rankings.
Run This Evaluation On Your Own Agency’s Reporting
This does not require a 6-month parallel test to apply. It requires pulling one report and checking it properly.
- Pull every referring domain from your current campaign’s report into Ahrefs or Semrush directly, rather than reading the agency’s summary.
- Check organic traffic for each linking page, not just its DR, and flag anything under 200 monthly sessions as low-impact regardless of the DR shown.
- Check whether editorial placements or niche edits make up most of your links, and if niche edits, whether the host pages have been updated recently.
- Ask your current agency directly for the traffic column on their next report. How they respond tells you most of what you need to know.
The Number On The Invoice Is Not The Number That Matters
None of the four agencies in this test were dishonest. They reported the metric the industry has trained clients to expect, Domain Rating, because it is the easiest number to put in a deck. The metric that actually correlated with ranking movement, organic traffic on the linking page, is harder to report favourably and rarely appears in a standard client-facing summary. A link building agency UK businesses can trust is the one willing to show that column without being asked twice.
This is the same evaluation framework RankFast applies to every link building engagement it runs, and the same standard held across every service RankFast offers: report the number that predicts the outcome, not the one that is easiest to sell.
Frequently Asked Questions
What did the 6-month agency comparison show?
The agency with the highest average linking-page traffic drove the largest ranking movement, and the agency with the lowest drove a negative one, despite a respectable average DR. At the four-agency level, price, DR, and traffic all moved together, so the traffic-specific finding comes from the underlying link-level analysis, not the agency averages alone.
How do you audit a link building agency before hiring them?
Request a sample report from a comparable client, then check each referring domain's traffic trend, topical relevance, anchor text diversity, and whether the same page links to multiple clients in similar niches. Reports showing only DR, with no traffic column, hide the metric that predicts results.
What is the difference between editorial links and niche edits in 2026?
Editorial links are placed within newly published content on active publishers. Niche edits insert a link into content published earlier. In this test, editorial links outperformed niche edits by 34 percent on ranking impact per link, likely tied to how quickly fresh content gets crawled.
