# SEO Data Bench

> Measured comparisons of SEO API output on fixed query cohorts. Published by the team behind Treg (treg.to); pages that evaluate Treg disclose it inline.

Every number is a call we made on a stated date against a declared cohort, with the raw response downloadable under /evidence. List prices are labelled as such.

## Benchmarks
- [Backlink counts across providers: same ranking, different numbers](https://seodatabench.com/benchmarks/backlink-overlap): For 20 domains on 15 September 2026, Serpstat, SE Ranking and Moz ranked the domains almost identically by referring domains (Spearman 0.97 to 0.98) while disagreeing on the counts by a median factor of 2.85 and up to 8.6× for one domain. Serpstat reported the lowest count for 19 of 20 domains, typically a third to a half of SE Ranking's or Moz's. Authority scores are on different scales and are not comparable: hunter.io is 60 on Serpstat, 90 on SE Ranking and 60 on Moz. Use any one provider consistently for relative comparisons; do not quote an absolute backlink count as a fact without naming the index.
- [Keyword volume coverage: which APIs report the long tail at all](https://seodatabench.com/benchmarks/keyword-data-coverage): On cohort v1 (50 keywords, US English, 15 September 2026) DataForSEO returned a volume for 31 of 50 keywords and matched Google Ads Keyword Planner exactly on every one of the 30 it shares, which is expected because its search_volume endpoint is Google Ads data resold. SE Ranking returned a volume for 25 of 50 and was within a factor of two of Google on 23 of 24, but reported 'serp api' at 3,100 against Google's 12,100 and had no value for several 10-to-30-a-month terms Google reports. Serpstat returned only 23 of the 50 keywords at all, ran about 20% higher than Google on the ones it had, and reported '1' for two terms Google shows no data for. None of the four reported a zero; the long tail comes back as null or absent, and the difference between those two is the coverage signal.
- [SEO API latency, failures and what a real workload costs](https://seodatabench.com/benchmarks/seo-api-latency): Across 314 calls on 15 September 2026, no route returned an error on a SERP or backlink call; the only failure was a DataForSEO bulk keyword call rate-limited because our own SERP calls to the same provider were running at the same time, which succeeded on retry. Stored-index routes answer in about a second (Serpstat SERP 1.0 s, all three backlink summaries 1.0 to 1.1 s at the median); live Google fetches take 1.5 s (SerpApi), 4.6 s (Cloro render), 7.9 s (DataForSEO) and 9.2 s (Treg routed, served by DataForSEO). A monthly workload of 200 keywords, 500 SERP checks and 50 backlink summaries costs between $0.55 and $8.60 depending on route, and about $1.40 on the cheapest live-data combination. The whole benchmark cost $1.99 at list prices.
- [Google SERP completeness: live fetch versus stored index](https://seodatabench.com/benchmarks/serp-completeness): Against a live rendered Google results page captured at the same time, DataForSEO's live SERP endpoint returned 95% of the real top-10 URLs and the same first result on 47 of 50 queries; SerpApi 90% and 47 of 50; Treg's routed SERP endpoint, which was served by DataForSEO on all 50 calls, 95% and 48 of 50. Serpstat's SERP is a stored index, not a fetch: it had no data for 30 of the 50 queries and, on the 20 it had, matched 70% of the live top 10 and the first result on 13. Cost per query ranged from $0.0005 (Serpstat) to $0.015 (SerpApi); median latency from 1.0 s (Serpstat, no fetch) and 1.5 s (SerpApi) to 7.9 s (DataForSEO live) and 9.2 s (Treg routed).

## Datasets
- [SEO API evaluation cohort v1: reproduce the benchmark](https://seodatabench.com/datasets/seo-api-evaluation-v1): Cohort v1 is 50 US-English buyer queries for data APIs and 20 vendor domains, frozen on 15 September 2026. It was sent on the same day through four keyword-volume routes (DataForSEO, Serpstat, SE Ranking, Google Ads Keyword Planner), five Google SERP routes (Cloro live render, DataForSEO, Serpstat, SerpApi, Treg routed) and three backlink routes (Serpstat, SE Ranking, Moz), one call per query or domain per route, one worker per provider. Every raw response is published as one JSON file per call. Re-running the whole cohort needs the Treg CLI and about $2.50 in API charges at list prices.

## Reference
- [Methodology](https://seodatabench.com/methodology): what the numbers mean, what they cannot show, and what could not be measured.
- [Evidence](https://seodatabench.com/evidence): raw responses, cohorts and runner scripts.
- [About](https://seodatabench.com/about): authors and the relationship to Treg.
- [Corrections](https://seodatabench.com/corrections): changes to published facts.