Where the coverage analyzer asks whether the right things are mentioned, this one asks what is mentioned at all and whether the markup confirms it. It is the more diagnostic of the two: it hands you the list a machine would build from your page, which is frequently not the list you thought you had written. Seeing the difference is usually more useful than the score.
Here is where that sits. The median SeoMods score is 79/100, which ranks it 5 of the 25 tools with a published benchmark by median. The ranking runs over the 25 tools on the main benchmark page; 1st is the highest median. The middle half of results falls between 21 and 79 — a band 58 points wide. That is wide: in this sample sites’ scores diverge markedly on this measure.
How the scores are distributed
An average on its own misleads. The chart below splits all 3,825 scans into five score bands, totalling 100%. The spread is wide and fairly even: in this sample sites land at very different levels on this measure.
- 0–19
- 20–39
- 40–59
- 60–79
- 80–100
| Score band | Entity Extractor |
|---|---|
| 0–19 | 20% · 756 |
| 20–39 | 12% · 449 |
| 40–59 | 17% · 656 |
| 60–79 | 31% · 1,176 |
| 80–100 | 21% · 788 |
| Total runs | 3,825 |
The grade mix
The same data as letters: A+ 95–100, A 85–94, B 75–84, C 65–74, D 50–64 and F below 50.
Where does my score sit?
The number reflects how many identifiable things the page names, how varied they are in type, and whether structured data pins them down. It is one of the few scores where reading the extracted list matters more than the figure — a page can score adequately while the extracted entities reveal it is about something subtly different from what was intended.
| Tool | P10 | P25 | Median | P75 | P90 | Mean | Lowest | Highest |
|---|---|---|---|---|---|---|---|---|
| Entity Extractor | 0 | 21 | 79 | 79 | 100 | 57 | 0 | 100 |
A score from the Entity Extractor can be read against the five values below. All of them are calculated from the scans in this benchmark: they place a SeoMods score within this sample, not in Google’s results.
| Percentile | Score | What it means for a score |
|---|---|---|
| P10 | 0 | About one scan in ten scored 0 or lower. Below it, a page is in the lowest tenth of this sample. |
| P25 | 21 | The edge of the bottom quarter. A score under 21 is below about three quarters of the scans. |
| Median | 79 | The middle of the sample: half the scans scored lower, half higher. A score of 79 is the typical result here. |
| P75 | 79 | Where the top quarter begins. A score of 79 or more is at or above about three quarters of the scans. |
| P90 | 100 | Where the top tenth begins. A score of 100 or more is among the highest 10% of scans in this sample. |
Some of these values coincide. That happens when many scans land on exactly the same score, so the bands between them are not separate bands at all — read the equal values as one step.
Percentiles describe the sites SeoMods scanned, not the whole web, and they apply to this tool only: the same number on another tool sits somewhere else.
The SeoMods checks most often unmet
The failures are about definition rather than quantity. Enough named entities, a mix of types rather than twenty instances of one, entities that recur rather than appear once, and structured data that says which specific thing an ambiguous name refers to. That last one is the difference between a page a machine can file and one it can only guess at.
| Check | Not met (share of scans) | Weight |
|---|---|---|
| Mix of entity types | 60% | Partial |
| Structured data to define entities | 52% | Partial |
| Enough named entities | 34% | Hard fail |
| Salient (repeated) entities | 33% | Partial |
The rate is the share of scans in which a SeoMods check was not met. It is SeoMods’ criterion, not Google’s assessment of these pages. “Hard fail” means the check’s full weight is deducted from the SeoMods score; “partial” means part of it.
The weakest areas
This tool groups its checks into categories. The chart below shows how often each category was the lowest-scoring one in a scan — which area is most often the weak link. It is not the average score in that category; it is how often it came last.
“Entity Richness” was the weakest area in 52% of scans.
What to fix first
Add structured data naming the main subject of the page and linking it to an authoritative identifier where one exists. That single addition resolves ambiguity that no amount of prose can, and it converts the page from something a machine infers a topic for into something that declares its own.
A typical scan passes 2.2 of 4 checks and leaves 1.8 open items behind. In this sample what brings a SeoMods score down is mostly not one large omission but accumulated small ones.
Run the Entity Extractor on your own page
Month by month
This page adds every published month together. Underneath it, each month is frozen as an edition of its own whose figures are never edited again — which is the only reason two months can be set side by side at all. Below is the Entity Extractor across the 2 published editions, with the combined figure in the last row of the table.
Between August 2026 and September 2026 the median went from 79 to 58, and the share of scans at 80 or above went from 24% to 19%. The two months do not measure the same sites — the September 2026 sample has about 3 times as many sites — so read the gap as a change in the cross-section measured, not as the same sites getting better or worse.
| Edition | Median | vs. previous | Scored 80+ | Issues/run | Runs |
|---|---|---|---|---|---|
| August 2026 | 79 | baseline | 24% | 1.6 | 981 |
| September 2026 | 58 | −21 pts | 19% | 1.9 | 2,844 |
| All published months, combined | 79 | — | 21% | 1.8 | 3,825 |
Reading a monthly difference. Each month is a separate sample, and its size changes from month to month. A change in sample size can move a median and a distribution by itself, so a monthly difference should not be mistaken for a change in scores: it does not show that sites improved or got worse.
Which checks moved
Each check’s rate in the first edition that measured it, beside its rate in the latest. Moves under two points are marked as no change: these are independent monthly samples of a shifting set of sites, and a difference that small is not evidence that anything on the web changed.
| Check | Not met in August 2026 | Not met in September 2026 | Change |
|---|---|---|---|
| Structured data to define entities | 43% · August 2026 | 55% · September 2026 | +12 pp |
| Salient (repeated) entities | 26% · August 2026 | 36% · September 2026 | +10 pp |
| Enough named entities Hard fail | 29% · August 2026 | 35% · September 2026 | +6 pp |
| Mix of entity types | 59% · August 2026 | 61% · September 2026 | no change |
What this score does not measure
- It is not a direct measurement of the signals every AI system uses. Which system uses which signal, and how, is not public; the score measures the checks SeoMods chose.
- It does not measure whether a page is actually cited in AI answers. It measures readiness on the page, not what any assistant returned.
- It is not a Google ranking. A high score does not guarantee a better position, and a low one does not mean Google has penalised the page.
About this benchmark. This benchmark is calculated from data on the sites SeoMods has scanned. The results do not necessarily represent all websites, or a random selection of them. The scores express SeoMods’ own measurement and scoring criteria; they are not ranking, quality, E-E-A-T or AI visibility scores issued by Google. Sample, anonymization and licence: the full methodology.
Frequently asked questions
Why does entity type variety matter?
Because a subject is normally made of several kinds of thing — a person at an organisation, in a place, working on a product, under a standard. A page naming only one type usually covers only one dimension of its subject. Variety is a proxy for having described the context rather than just the thing.
Do I need Wikidata or schema.org identifiers?
They are not required, but they are the strongest available way to remove ambiguity. A sameAs pointing at an authoritative record turns "we mean this specific company" from an inference into a statement. For any name shared by more than one well-known thing, it is worth the few minutes.
How does this differ from the Entity Coverage Analyzer?
This one reports what is there; the coverage analyzer judges whether what is there is enough for the subject. Use this to check that a machine reads your page as being about what you think it is about, and the coverage analyzer to find what you left out.
What is a good score on the Entity Extractor?
In this dataset the median SeoMods score is 79/100, so scoring 79 sits exactly in the middle of the sample. 79 is the start of the upper quartile and 100 or above is the top 10%; below 21 is the bottom quarter. Those thresholds are specific to this tool — the 25 tools with a published benchmark have very different medians, so 70 here and 70 elsewhere do not mean the same thing. They place a page among the sites SeoMods scanned; they say nothing about Google rankings.
Is the score from the Entity Extractor issued by Google?
No. It is a benchmark score calculated against the check list SeoMods defined for this tool. A SeoMods check going unmet does not mean Google has penalised the page or ranked it lower.
Which SeoMods check goes unmet most often on the Entity Extractor?
“Mix of entity types”, which was not met in 60% of scans. The largest losses on the SeoMods score, though, come from the checks scored as hard fails — this tool has 1, the most common being “Enough named entities” at 34%.
Questions that govern the whole series — how the sample is collected, how sites are anonymized, how often the figures are refreshed — are answered in the methodology on the hub.
Using this data
The figures, tables and charts on this page are free to reuse under CC BY 4.0; the only condition is a link back to this page. Suggested citation: SeoMods — Entity Extractor benchmarks, 2 October 2026, https://seomods.com/seo-score-benchmarks/entity-extractor. If you need to cite by date, use the September 2026 edition instead: this page changes at every refresh and that one never does.