Where the gap finder asks which questions are missing, this asks whether the ones that are there are in usable shape. It is the stricter of the two and the scores show it — a page can have a perfectly good FAQ section written as prose and still fail nearly every check here, because none of it is marked as question and answer. The content exists; the structure does not.
Here is where that sits. The median is 30/100, which ranks it 22 of the 25 measures in this dataset. The middle half of results falls between 30 and 30 — a band 0 points wide. That is narrow: sites resemble each other on this measure, which usually means a platform is doing most of the work.
How the scores are distributed
An average on its own misleads. The chart below splits all 1,481 scans into five score bands, totalling 100%. The distribution banks up against the floor: most pages score badly, so even a modest result here puts you ahead of the majority.
- 0–19
- 20–39
- 40–59
- 60–79
- 80–100
| Score band | AI FAQ |
|---|---|
| 0–19 | 13% · 197 |
| 20–39 | 72% · 1,069 |
| 40–59 | 8% · 115 |
| 60–79 | 6% · 84 |
| 80–100 | 1% · 16 |
| Total runs | 1,481 |
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 your score stands
The floor here is low enough that the score is best read as a checklist state rather than a grade. Either the page has recognisable question-and-answer structure or it does not, and moving from "does not" to "does" is a formatting exercise on content you have already written. Very few pages sit meaningfully in between.
| Tool | 10th | 25th | Median | 75th | 90th | Mean | Lowest | Highest |
|---|---|---|---|---|---|---|---|---|
| AI FAQ | 0 | 30 | 30 | 30 | 40 | 29.7 | 0 | 100 |
How to read it: 30 is bottom-quartile, 30 is a typical page, 30 starts the upper quartile and 40 or above is the top 10%. These thresholds apply to this tool only.
The checks that fail most
The checks want three things aligned: a question phrased as one, an answer immediately below it and short enough to be lifted whole, and FAQPage or QAPage markup declaring which is which. Any two without the third leaves an engine inferring, and it usually declines to.
| Check | Did not pass | Weight |
|---|
“Hard fail” means the check’s full weighting is lost; “partial” means only part of the score is. Rates are the share of scans in which the check ran and did not pass.
What to fix first
Convert the questions you already answer into headings phrased as questions, then generate the matching FAQ schema from them rather than writing it separately — schema that drifts out of step with the visible text is worse than none, because it makes a claim the page contradicts.
A typical scan passes 0.0 of 0 checks and leaves 0.0 open items behind. What drags a score down is usually not one catastrophe but accumulated small omissions.
Run the AI FAQ Optimizer on your own page
Month by month
The figures on this page are replaced at every refresh. At the end of each month a permanent edition is published for that month and never edited again — which is the only reason this series can be compared at all. One edition has been published so far; when the second lands this section will set the two months side by side.
| Edition | Median | vs. previous | Scored 80+ | Issues/run | Runs |
|---|---|---|---|---|---|
| August 2026 | 30 | baseline | 1% | 0.0 | 1,481 |
Frequently asked questions
Do the questions have to match how people search?
They should be recognisably the same question, in natural phrasing. Matching a query word for word is neither necessary nor a good idea — engines handle paraphrase well, and a heading contorted to include an exact search string reads badly to the human it is meant to help.
Can FAQ schema be added to any page?
Technically yes, and it should not be. The markup asserts that the page contains those questions and answers as visible content; if it does not, the assertion is false and treated as such. Add the schema to pages that genuinely answer the questions, and it is a description rather than a claim.
Should the answer be short or thorough?
Both, in that order. Two to four sentences that answer completely, then any depth underneath. A short answer that omits the qualification is not extractable either, because an engine will not repeat something it cannot bound — completeness is the requirement and brevity is the constraint.
What is a good score on the AI FAQ Optimizer?
In this dataset the median is 30/100, so scoring 30 puts you exactly in the middle. 30 is the start of the upper quartile and 40 or above is the top 10%; below 30 is the bottom quarter. Those thresholds are specific to this tool — the 25 measures in this dataset have very different medians, so 70 here and 70 elsewhere do not mean the same thing.
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 — AI FAQ Optimizer benchmarks, 2 September 2026, https://seomods.com/seo-score-benchmarks/ai-faq-optimizer. If you need to cite by date, use the August 2026 edition instead: this page changes at every refresh and that one never does.