Research methodology · Reviewed 23 July 2026
How the OpptiAI image SEO audit measures WordPress sites
The definitions, inclusion rules, privacy protections, and limitations behind the anonymised aggregates used in the State of WordPress Image SEO report.
Definitions
- Eligible site
- A distinct public domain with a completed audit, at least one image scanned, and a successful crawl status.
- Sample unit
- Distinct domains, not audit runs. Repeating an audit on one domain does not increase the sample size.
- Missing alt text
- An in-scope image whose rendered alt attribute is absent or empty when the image is not identified as decorative.
- Weak alt text
- An in-scope image whose current text falls below the audit’s documented quality threshold. This is a prioritisation signal, not a WCAG conformance verdict.
- Coverage
- The share of in-scope crawled images that have useful alt text under the audit heuristic.
- Publication threshold
- Aggregate percentages remain hidden until at least 25 eligible distinct domains are available.
Privacy
Public aggregates never expose email addresses, submitted site URLs, normalised domains, priority-page lists, or sample image URLs. The public snapshot contains only sample size, aggregate measures, issue labels, threshold status, and timestamps.
Limitations
Sites are self-selected users of a free public audit, not a random sample of all WordPress sites. Crawlability, lazy loading, blocked assets, and page selection can affect what is observed. Scores are heuristic prioritisation aids, not legal accessibility certification or guaranteed ranking predictions.
Reproduce the public status
The current anonymised snapshot is available at /api/image-seo-stats. Read the State of WordPress Image SEO 2026 for the public report, or run the audit on your own site.
FAQ
Is the image SEO audit a WCAG certification?
No. Missing and weak alt signals are heuristic prioritisation aids. They are not a legal accessibility certification or a whole-site WCAG conformance verdict.
Why do aggregate percentages stay hidden until 25 domains?
Small self-selected samples overstate precision. The publication threshold waits for at least 25 eligible distinct domains before public aggregates appear on the State report.
Do repeat audits on the same site increase the sample size?
No. The sample unit is distinct domains, not audit runs. Re-crawling one domain does not inflate N.
What personal data appears in the public stats API?
None of the site-identifying fields. Public snapshots include sample size, aggregate measures, issue labels, threshold status, and timestamps — not emails, URLs, or sample image paths.