Data study · 26 Oct 2026
How often CitedBy findings are wrong: our own sites, read one by one
We read 66 findings on our own sites against their pages, in two readings: 3 were wrong. First reading: we ran the current audit engine on saved crawls of hexaciv.com, civiqpass.com and citedby.eu (taken 23 Oct 2026). It raised 48 findings. We read 36 of them against the page they came from: 0 were wrong, 2 partly right and 34 right. The other 12 we did not judge, because a replay cannot fetch robots.txt, llms.txt, the live sitemap, every response header or /favicon.ico.
Reading 1, checked 26 Oct 2026 · three of our own sites · 48 findings · Download the per-finding CSV
The numbers
| Site | Findings | Judged | Right | Partly right | Wrong | Not judged |
|---|---|---|---|---|---|---|
| citedby.eu | 19 | 15 | 13 | 2 | 0 | 4 |
| hexaciv.com | 17 | 12 | 12 | 0 | 0 | 5 |
| civiqpass.com | 12 | 9 | 9 | 0 | 0 | 3 |
| All three | 48 | 36 | 34 | 2 | 0 | 12 |
A finding is one check that fired on a site, with its list of affected pages (for example "title or description is cut off, 1,556 pages"). The rate is wrong findings over judged findings. Findings we could not judge are left out of both numbers and listed below.
Per check
All 78 findings the two readings raised (66 judged), counted by the check that raised them: 36 checks. Most checks fired once or twice on our four sites, so one check's n is small and its wrong count says little by itself. Customer marks fill the live per-check rates at /api/precision, which publishes a rate from 10 reviewed findings.
| Check | Findings read | Judged | Right | Partly right | Wrong | Not judged |
|---|---|---|---|---|---|---|
author_anonymous | 6 | 6 | 5 | 1 | 0 | 0 |
internal_param_links | 6 | 6 | 6 | 0 | 0 | 0 |
content_duplicate | 4 | 4 | 3 | 0 | 1 | 0 |
crawl_refused | 4 | 2 | 1 | 1 | 0 | 2 |
deep_page | 4 | 4 | 4 | 0 | 0 | 0 |
faq_schema_widespread | 4 | 4 | 4 | 0 | 0 | 0 |
noindex | 4 | 4 | 4 | 0 | 0 | 0 |
entity_sameas_missing | 3 | 3 | 3 | 0 | 0 | 0 |
link_wrong_language | 3 | 3 | 2 | 0 | 1 | 0 |
llms_full_missing | 3 | 0 | 0 | 0 | 0 | 3 |
llms_txt_missing | 3 | 0 | 0 | 0 | 0 | 3 |
sitemap_lastmod | 3 | 2 | 2 | 0 | 0 | 1 |
snippet_cut | 3 | 3 | 3 | 0 | 0 | 0 |
faq_not_visible | 2 | 2 | 2 | 0 | 0 | 0 |
no_validators | 2 | 2 | 2 | 0 | 0 | 0 |
outbound_moved | 2 | 2 | 2 | 0 | 0 | 0 |
public_not_cacheable | 2 | 2 | 2 | 0 | 0 | 0 |
render_blocking | 2 | 2 | 2 | 0 | 0 | 0 |
canonical_noindex | 1 | 1 | 1 | 0 | 0 | 0 |
csp_unsafe_inline | 1 | 1 | 1 | 0 | 0 | 0 |
description_length | 1 | 1 | 1 | 0 | 0 | 0 |
favicon_missing | 1 | 0 | 0 | 0 | 0 | 1 |
heavy_page | 1 | 1 | 1 | 0 | 0 | 0 |
hreflang_not_reciprocal | 1 | 1 | 1 | 0 | 0 | 0 |
hreflang_x_default_missing | 1 | 1 | 1 | 0 | 0 | 0 |
hsts_missing | 1 | 0 | 0 | 0 | 0 | 1 |
internal_hash_links | 1 | 1 | 1 | 0 | 0 | 0 |
internal_redirect_link | 1 | 1 | 1 | 0 | 0 | 0 |
js_changes_page | 1 | 1 | 1 | 0 | 0 | 0 |
low_text_ratio | 1 | 1 | 1 | 0 | 0 | 0 |
no_main_landmark | 1 | 1 | 1 | 0 | 0 | 0 |
schema_article_props | 1 | 1 | 1 | 0 | 0 | 0 |
schema_faq_missing | 1 | 1 | 0 | 0 | 1 | 0 |
security_headers_missing | 1 | 0 | 0 | 0 | 0 | 1 |
title_brand_first | 1 | 1 | 1 | 0 | 0 | 0 |
title_length | 1 | 1 | 1 | 0 | 0 | 0 |
Second reading: our latest live audits
The first reading used saved crawls, so it could not see pages published after 23 Oct 2026 or a site with no saved crawl. For the second we took the latest finished audit of hexaciv.com, civiqpass.com, citedby.eu and resaleiq.dev (24 to 26 Oct 2026, the engine as it was that day): 30 findings. Each was checked against the live page. 16 were right on the live page, 11 were right when the audit ran and describe pages that have changed since, and 3 were wrong.
| Site | Findings | Right now | Right then, page changed | Wrong |
|---|---|---|---|---|
| civiqpass.com | 4 | 1 | 1 | 2 |
| resaleiq.dev | 14 | 10 | 3 | 1 |
| citedby.eu | 6 | 2 | 4 | 0 |
| hexaciv.com | 6 | 3 | 3 | 0 |
| All four | 30 | 16 | 11 | 3 |
Wrong: civiqpass.com content_duplicate; civiqpass.com link_wrong_language; resaleiq.dev schema_faq_missing.
| Site | Check | Verdict | Pages | What the page shows | How checked |
|---|---|---|---|---|---|
| civiqpass.com | content_duplicate | wrong | 4 | The named China pair has 4,917 and 1,911 words and a 4-gram similarity of 0.21; the India pair 0.31. They share quoted official wording but are not near-duplicates. | live |
| civiqpass.com | link_wrong_language | wrong | 78 | The first sample page links /contact (English) and also /ar/contact (200, lang ar), but the report says /contact has no Arabic version. The Arabic page exists and the same page links it, though /contact does not list it in hreflang. | stored crawl + live |
| resaleiq.dev | schema_faq_missing | wrong | 2 | The question headings on /blog are post titles in a list, not questions with answers. The rule was fixed on 26 Oct 09:10 (RES-4753), after the audit. | live + code |
| citedby.eu | author_anonymous | right | 328 | JSON-LD author on /ai-visibility-audit and /blog/ai-crawler-check-in-ci is an Organization ('CitedBy' or 'CitedBy editorial team'); no named person on any of the three sample pages. | live |
| citedby.eu | crawl_refused | right, page changed since | 6 | Six free-tool example pages answered 429 to the audit crawl (RES-4854 records that our crawl was refused). All three samples answer 200 today at a lower request rate. | change record |
| citedby.eu | hreflang_not_reciprocal | right, page changed since | 8 | At audit the blog pages listed an nl alternate that is noindex. RES-4854 (site, 26 Oct, after the audit) stopped listing noindex alternates; the samples no longer list one. | change record + live |
| citedby.eu | internal_param_links | right, page changed since | 388 | At audit the data pages linked /signin?lang=fr&next=... RES-4854 (26 Oct, after the audit) moved the sign-in link to a language path; the lang parameter is gone. | change record + live |
| citedby.eu | internal_redirect_link | right, page changed since | 30 | At audit /backlinks linked /backlinks/www.gov.uk, which redirects. RES-4854 (26 Oct, after the audit) made those links skip the www redirect. | change record + live |
| citedby.eu | noindex | right | 4 | /de/sample, /es/sample, /fr/sample each send 'x-robots-tag: noindex' and <meta name=robots content=noindex>; English /sample is index,follow (translated copies of the sample report kept out on purpose). | live |
| civiqpass.com | canonical_noindex | right, page changed since | 4 | The stored crawl shows /fr/pack/lost and /en/pack/lost with a canonical pointing at /pack/lost, which is noindex. Today each is canonical to itself. | stored crawl + live |
| civiqpass.com | internal_param_links | right | 7 | All 3 samples link to /es/?start=mock&hop=blog_cta (or hop=civiqpass_seo_try); the example in the report said src=blog_cta, so the parameter name changed. Internal link count 19 on page 1 matches. The tagged URL is noindex,follow with canonical /es/, so a notice is fair. | live |
| hexaciv.com | author_anonymous | right, page changed since | 19 | The stored crawl shows Article markup with no author on the samples at audit time. Today the same pages carry an Organization author. | stored crawl + live |
| hexaciv.com | deep_page | right, page changed since | 3 | The stored crawl gave /ar/partners and /en/formateurs a click depth of 4, reached only from /ar/embed and /en/embed (depth 3). Today they are also linked from /a-propos pages at depth 2, so they are now 3 clicks away. | stored crawl + live |
| hexaciv.com | faq_schema_widespread | right | 1 | Sampled (12 per sitemap section, 12 of 219 fiches): FAQPage JSON-LD on 12/12 question and fiche pages, 6/12 core, 2/12 centres; weighted estimate about 2300 of ~2508 pages (92%) vs reported 2367 of 2515 (94%). Not a full count; the >50% claim is certainly true. Unclear if the markup on every page is warranted. | live |
| hexaciv.com | js_changes_page | right | 1 | /ar/pass-examen-civique: ~350 words in raw HTML (report 359), 868 words in headless Chromium innerText (report 844), same URL and title; a script adds FAQ/answer sections AI crawlers will not read. | live |
| hexaciv.com | noindex | right | 88 | All 3 samples (/ar/centres/91-essonne, /ar/contact, /ar/legal) return 200 with 'x-robots-tag: noindex'. contact and legal are meant to be private/kept out; the /ar centres page is full Arabic content while /en and root /centres/91-essonne are indexable, so whether that one is deliberate is not clear. | live |
| hexaciv.com | outbound_moved | right, page changed since | 657 | At audit the sample pages linked the Etalab licence address, which answers 301 to https://www.data.gouv.fr (read live today). The pages now link data.gouv.fr/pages/legal/licences/etalab-2.0, so the finding no longer applies. | live redirect + page since changed |
| resaleiq.dev | author_anonymous | right | 9 | All 3 /compare pages carry Article JSON-LD with author Organization 'Resale IQ' (url /about/data-team); no named person. | live |
| resaleiq.dev | content_duplicate | right | 41 | bags/caps/jeans: same template, word-sequence similarity 0.89 (bags/caps), 0.85 (bags/jeans), 0.84 (caps/jeans); only the category word, top brand and prices differ. 'Nearly the same text' holds by wording, though each page has its own data and title. | live |
| resaleiq.dev | csp_unsafe_inline | right | 1 | Header on https://resaleiq.dev/: content-security-policy ... script-src 'self' 'unsafe-inline' https://www.googletagmanager.com, no nonce or sha hash. | live |
| resaleiq.dev | deep_page | right | 21 | /buy/acne-studios, /buy/bershka, /buy/corteiz are not linked from home, from any of the 89 pages home links to, or from any of the 697 pages those link to (outlinks of all fetched), so they are at least 4 clicks deep. Exactly 4 not separately measured. | live |
| resaleiq.dev | description_length | right, page changed since | 3 | The stored description of /de/accuracy was 237 characters at audit. The sample pages are 153 to 157 characters today. | stored crawl + live |
| resaleiq.dev | faq_schema_widespread | right | 1 | The 7 sitemaps list 377 URLs; I fetched all 377 and 337 carry FAQPage JSON-LD (89.4%), vs reported 335 of 375. | live |
| resaleiq.dev | internal_param_links | right | 291 | The docs page links one internal address with a src tag (12 internal links by our own count today against 11 in the report); the two blog samples carry 4 to 6 utm or src tagged links each. All tagged copies checked are noindex,follow with a canonical to the clean URL, so 'warning' overstates it. The site-wide count of 805 was not checked. | live |
| resaleiq.dev | link_wrong_language | right, page changed since | 416 | The stored crawl shows /de/accuracy linking /buy, and /de/buy answers 307 to the English /buy. The link is no longer on the page today. | stored crawl + live |
| resaleiq.dev | low_text_ratio | right | 239 | Text excluding nav/header/footer is 3.6% / 3.0% / 4.1% of the HTML for / , /about, /about/data-team (report 3.6% for /); inline scripts are ~55% of the HTML. Counting nav and footer text too gives 6.8-8.1%, so the number depends on the definition. | live |
| resaleiq.dev | noindex | right | 9 | /de/founding and /es/founding: <meta robots noindex, nofollow>; /de/legal (Impressum) has the meta plus x-robots-tag noindex, nofollow, noarchive. Legal and offer pages look intentionally kept out. | live |
| resaleiq.dev | public_not_cacheable | right | 36 | /tools, /tools/bulk-listings, /tools/bundle-calculator all send 'cache-control: private, max-age=0, must-revalidate' (while /about sends public, s-maxage=300). | live |
| resaleiq.dev | render_blocking | right | 32 | /, /de, /de/tools each have 4 <link rel=stylesheet> in <head> with no media/onload and no blocking script (the only sync-looking src script is nomodule); matches '0 scripts, 4 stylesheets'. | live |
| resaleiq.dev | title_length | right, page changed since | 1 | The stored title of /de/accuracy was 66 characters at audit ('... von Resale IQ'); it is 52 characters today. | stored crawl + live |
Download the second CSV. A finding that was right when the audit ran counts as right even if the owner has fixed the page since; it is shown separately so you can see how many.
Before this check
- MeasuredOn 23 Oct 2026 the previous version of the engine raised 41 findings on four of our sites (these three and resaleiq.dev). Reading them by hand found 5 false alarms and 5 findings that could no longer be reproduced in the page.
- MeasuredThe checks behind those false alarms were changed on 24 and 25 Oct 2026. This replay is the first read of the current engine.
- Not measuredresaleiq.dev has no saved crawl, so it is not in the 48 (it is in the second reading). The checks that need a model key or a browser did not run in the replay, so this table does not cover them.
Every finding with its verdict
| Site | Check | Verdict | Pages | What the page shows | Still true 26 Oct |
|---|---|---|---|---|---|
| citedby.eu | link_wrong_language | hit | 146 | Snapshot: the es home links /methodology; no /es/methodology page in the crawl and /methodology had no hreflang alternates. Fixed since: /es/methodology exists (RES-4518). | no |
| citedby.eu | crawl_refused | partly | 23 | Snapshot: 19 pages answered HTTP 429. The 4 'no answer' rows are /signin?lang= URLs the snapshot does not hold. The 19 refusals are real; the 4 no-answer rows are a snapshot gap, so the count is overstated by 4. | not checked |
| citedby.eu | snippet_cut | hit | 16 | Meta description of /terms, /refund and /privacy ends with '...'. | not checked |
| citedby.eu | entity_sameas_missing | hit | 1 | Home JSON-LD: sameAs is [https://citedby.eu/about, https://citedby.eu/methodology]. | not checked |
| citedby.eu | faq_not_visible | hit | 1 | FAQPage JSON-LD names 'What is an AI visibility audit?' and 'What does the CitedBy audit cost?'; the visible headings were different questions. Fixed since: the visible questions now match the markup. | no |
| citedby.eu | author_anonymous | partly | 79 | The /sample pages carry a Report (an Article subtype) node with no author. Blog and guide pages carry an Article whose author is the Organization 'CitedBy editorial team', which the check names as anonymous (6 random pages checked). Of the 79 flagged pages, 5 (/sample x5) have no author and 74 name an organization (73 'CitedBy editorial team', 1 'CitedBy'), so the detail line is exact for 5 and loose for 74. Second read changed this from right to partly right. | not checked |
| citedby.eu | public_not_cacheable | hit | 30 | Snapshot headers of /plans and /fr/plans: Cache-Control no-store, status 200. Fixed since: /plans is public, max-age 300. | no |
| citedby.eu | deep_page | hit | 8 | Shortest link path in the snapshot from the home page is 4 clicks. | not checked |
| citedby.eu | heavy_page | hit | 5 | Snapshot: /sample HTML is 622,525 bytes. | not checked |
| citedby.eu | no_main_landmark | hit | 5 | Snapshot HTML of /sample has no <main>; landmarks are header and section only. | not checked |
| citedby.eu | internal_param_links | hit | 4 | The /fr/plans HTML contains href="/signin?lang=fr&next=%2Fapp.html". | not checked |
| citedby.eu | outbound_moved | hit | 4 | Fetched 26 Oct: the linked help article answers 301 to a different address on the vendor's new help domain. | yes |
| citedby.eu | faq_schema_widespread | hit | 1 | Snapshot: 138 of 221 indexable HTML pages carry FAQPage JSON-LD (62%, over the half threshold). The engine's 162 is a sampled subset; the share agrees. | not checked |
| citedby.eu | no_validators | hit | 1 | Snapshot headers: 0 of 253 HTML pages have an ETag or Last-Modified. | not checked |
| citedby.eu | schema_article_props | hit | 1 | The Article JSON-LD of the snapshot page has no datePublished. Fixed since: the live page has datePublished. | no |
| citedby.eu | favicon_missing | not judged | 1 | The home page has no <link rel=icon>, but /favicon.ico exists (HTTP 200 on 26 Oct, in the repository since 16 Oct). The replay cannot probe /favicon.ico. Replay artifact: the probe is not replayable. | no |
| citedby.eu | llms_txt_missing | not judged | 1 | The snapshot holds no llms.txt (bodies of robots.txt and llms.txt were not saved); /llms.txt answers 200 on 26 Oct. Replay artifact. | no |
| citedby.eu | llms_full_missing | not judged | 1 | The snapshot holds no llms-full.txt; /llms-full.txt answers 200 on 26 Oct. Replay artifact. | no |
| citedby.eu | sitemap_lastmod | not judged | 1 | The snapshot's sitemap was rebuilt from the list of addresses without dates; the live sitemap has 453 lastmod values. Replay artifact. | no |
| hexaciv.com | internal_hash_links | hit | 2226 | The /en/ HTML holds 11 hrefs of the form #/ and #/account, #/exam/1. | not checked |
| hexaciv.com | internal_param_links | hit | 1683 | The /en/ HTML holds 6 hrefs with ?lang=en (/legal?lang=en, /terms?lang=en, /privacy?lang=en, /refund?lang=en, /contact?lang=en, /tcf-irn-tef-irn?lang=en). | not checked |
| hexaciv.com | snippet_cut | hit | 1030 | Checked all 1,030 flagged pages: every meta description ends in '…' or '...'. | not checked |
| hexaciv.com | content_duplicate | hit | 97 | Text of the two pages is 87% identical; 8 random flagged pages share 69-90% of their 4-word phrases with their nearest flagged page. The pages differ in place names, counts and the centre list. Templated pages: the statement is true, whether it matters is the owner's call. | not checked |
| hexaciv.com | faq_not_visible | hit | 3 | Checked the fr, en and ar pages: neither JSON-LD question appears in the visible text. | not checked |
| hexaciv.com | entity_sameas_missing | hit | 1 | Home HTML has an Organization node and no sameAs. | not checked |
| hexaciv.com | author_anonymous | hit | 19 | 5 random flagged pages: each has an Article node with no author. | not checked |
| hexaciv.com | noindex | hit | 5 | Snapshot headers: /contact, /en/guides, /privacy, /refund and /terms answer X-Robots-Tag: noindex. A notice of a fact; four of the five are legal pages that are noindex on purpose. | not checked |
| hexaciv.com | deep_page | hit | 1 | Shortest link path in the snapshot is 4 clicks; it is the only page at depth 4. | not checked |
| hexaciv.com | faq_schema_widespread | hit | 1 | Snapshot: 2,067 of 2,235 HTML pages carry FAQPage JSON-LD (92%); the report's 91.5% agrees. | not checked |
| hexaciv.com | no_validators | hit | 1 | Snapshot: 0 of 2,235 pages recorded an ETag or Last-Modified. Fixed since: the home page sends an ETag on 26 Oct. | no |
| hexaciv.com | sitemap_lastmod | hit | 1 | The saved sitemap: 1,848 URLs, 1,802 dated, 1,772 of them 2026-10-21. The live sitemap is an index of 8 files now, with different dates. | no |
| hexaciv.com | crawl_refused | not judged | 29 | The snapshot holds none of the 29 URLs (/legal, ?lang= pages); /legal answers 200 on 26 Oct. Replay artifact. | no |
| hexaciv.com | hsts_missing | not judged | 1 | The snapshot keeps only some response headers, not HSTS. The live home page sends HSTS on 26 Oct. Replay artifact. | no |
| hexaciv.com | security_headers_missing | not judged | 1 | The snapshot keeps only some response headers. The live home page sends all four on 26 Oct. Replay artifact. | no |
| hexaciv.com | llms_txt_missing | not judged | 1 | The snapshot holds no llms.txt; /llms.txt answers 200 (22,839 bytes) on 26 Oct. Replay artifact. | no |
| hexaciv.com | llms_full_missing | not judged | 1 | The snapshot holds no llms-full.txt; /llms-full.txt answers 200 on 26 Oct. Replay artifact. | no |
| civiqpass.com | snippet_cut | hit | 1556 | Checked every flagged page: 1,399 meta descriptions end in '…' or '...'; the other 157 have a title cut mid-phrase with '…' (for example 'CiviqPass | El Museo Reina Sofía conserva el óleo… · CCSE 4025'). | not checked |
| civiqpass.com | internal_param_links | hit | 1516 | The home HTML holds hrefs /?lang=en, /about?lang=en, /legal?lang=en, /terms?lang=en and more; 30 random flagged pages each hold at least one such link. | not checked |
| civiqpass.com | content_duplicate | hit | 17 | Main text of the flagged question pages overlaps 58-89% in 4-word phrases with their nearest flagged page; two CCSE pages that ask the same thing in different words have 84% of their text in common. Different questions that quote the same source text: the statement is true, whether it matters is the owner's call. | not checked |
| civiqpass.com | entity_sameas_missing | hit | 1 | Home JSON-LD: an Organization node with no sameAs. | not checked |
| civiqpass.com | author_anonymous | hit | 1760 | 6 random flagged pages: each has an Article node with no author. | not checked |
| civiqpass.com | title_brand_first | hit | 988 | All 988 flagged titles start with 'CiviqPass' and are 55 characters or longer (for example 'CiviqPass | Preguntas CCSE sobre la Constitución con respuesta', 62). | not checked |
| civiqpass.com | hreflang_x_default_missing | hit | 6 | All 6 flagged pages list hreflang fr, ar and es and no x-default. | not checked |
| civiqpass.com | render_blocking | hit | 1 | The home <head> holds two scripts without defer or async (lang, cq-theme) and one stylesheet (/cq.css). | not checked |
| civiqpass.com | sitemap_lastmod | hit | 1 | The saved sitemap rows: 1,898 URLs, all dated, 1,595 of them 2026-10-22. | not checked |
| civiqpass.com | crawl_refused | not judged | 44 | The snapshot holds none of the 44 URLs (/legal, /terms, /privacy and ?lang= pages); /legal answers 200 on 26 Oct. Replay artifact. | no |
| civiqpass.com | llms_txt_missing | not judged | 1 | The snapshot holds no llms.txt; /llms.txt answers 200 (64,294 bytes) on 26 Oct. Replay artifact. | no |
| civiqpass.com | llms_full_missing | not judged | 1 | The snapshot holds no llms-full.txt; /llms-full.txt answers 200 on 26 Oct. Replay artifact. | no |
The CSV also holds what the report said, an example URL and a note for each finding.
Method and limits
- Replay: the audit code that runs for customers, run on pages saved on 23 Oct 2026, with no network access. A request for a page that was not saved fails. The replay cannot fetch robots.txt, llms.txt, the real sitemap file, the response headers hexaciv.com's snapshot did not keep, or probe /favicon.ico. Findings that depend on those are marked "not judged".
- Verdicts: right means the statement is true of the page as saved; partly right means it is true for some pages or a number in it is off; wrong means it is untrue of the page. For findings that list many pages we read every flagged page where a script could (descriptions that end in an ellipsis, titles that start with the brand, links with a language parameter) and a random sample of 5 to 30 pages for the rest.
- Still true 26 Oct: a fetch on 26 Oct 2026 of the live page or file. "No" means the site has been fixed since the crawl; it does not make the finding wrong.
- What a right finding does not say: that the owner has to act on it. Pages built from one template, such as the question pages on civiqpass.com and the centre pages on hexaciv.com, are flagged as near-duplicates because their text is.
- Second reading: the audits are the ones our own scheduled runs produced for the four sites (not replays). Each finding was checked against the live page with our crawler,
CitedBot/1.0. A verdict that disagreed with the report was checked again against the page as stored at audit time. Right then, page changed means the stored crawl shows what the report said and the live page no longer does; those are counted as right. - Who judged: CitedBy read all findings. A second reader then re-read 15 of the judged findings (6 on citedby.eu, 3 on civiqpass.com, 6 on hexaciv.com) from the raw saved pages, counting each statement without our notes: 0 wrong, 2 partly right. One verdict changed from right to partly right (citedby.eu
author_anonymous) and is the one in the table. Findings from customer audits are counted separately, from the "This is wrong" button and from fixes you verify, in How often we are wrong.
See the same read on your site
We crawl your site up to your plan's page limit and give you the exact fix for each issue. From €9, 14-day money-back guarantee.
Checked on 26 Oct 2026 · Source: Data (CSV) · Method: How CitedBy checks · Change log: What changed