Wikipedia Risk Report 2026: Rejection, Deletion, Sources

How risky is it to get a Wikipedia page in 2026?

Risky enough to assess first: roughly two in three new drafts in the public sample were declined, and the risk is largely predictable from the source list. A page is achievable for a minority of subjects, no provider can guarantee a community outcome, and the one input you control before spending is the quality of independent coverage.

The AfC figures come from Lumino Digital, a third-party agency whose published analysis we cite as a benchmark, not as our own measurement:

Read these as the risk among subjects that tried. Public samples skew toward English Wikipedia and toward applicants who submitted, and they under-represent the many companies that never apply because they already know they would not qualify. We have not verified the sample independently; we have not found a comparable public dataset to cross-check it.

What do we base this report on?

On public, citable evidence, plus our own assessments clearly labelled as ours. We did not invent statistics, name clients or present third-party numbers as our measurements.

• Third-party benchmark: the Lumino Digital figures above, attributed each time.
• Wikipedia's own process pages: Articles for creation, Articles for deletion (a discussion runs for at least seven days before a decision based on consensus), the deletion policy and speedy deletion.
• Public decline reasons: AfC decline notices are templated and public. The recurring themes are notability, independent sourcing and promotional tone; see why drafts get rejected.
• Our notability index: the percentage of WikiBusines client assessments between 2018 and 2026 rated low, medium or high notability confidence, by industry. This is our own assessment of our own caseload, not a random sample of Wikipedia, and it is not audited.

Which gates stop a page?

Two, and they fail for overlapping reasons. Gate one is AfC, the volunteer review queue that new drafts from editors with a conflict of interest go through. Gate two is deletion: even a published article can be nominated at Articles for Deletion (AfD) or, if blatantly promotional, removed through speedy deletion without a discussion.

The notability guideline for companies sets the tone: only a small percentage of organizations meet the requirements for an article (WP:NCORP). A decline is therefore normal and one decline is not fatal, but a rushed second attempt on the same sources usually fails the same way. Budget months for review, not days, and expect a deletion nomination to be survivable only with strong sources. For what sinks pages after publication, see why pages get deleted.

How does risk differ by industry?

Sharply. Our notability index covers 13 industries, with the share of our client assessments rated low, medium and high notability confidence. These are our assessments of cases that came to us, so they describe our caseload.

The pattern fits the public benchmark: where coverage is mostly promotional or sits in thin trade outlets, assessments come out low; where independent journalism and criticism are abundant, they come out higher. Very few cases anywhere reach high confidence. Full methodology and downloads are on the notability index page; industry guidance is on our pages for fintech, crypto, tech companies and healthcare.

Which sources help and which hurt?

Independent, reliable, in-depth secondary coverage helps; anything the subject can manufacture does not. If you read enough public declines, the hierarchy becomes obvious.

Two rules cause most disappointments. Press is not independence: a feature that began as your PR pitch, ran as sponsored content or reprinted your announcement is not independent coverage, whatever the masthead. Volume is not depth: the general notability guideline asks for sources that address the topic directly and in detail, so three strong independent profiles beat fifty name-drops. Strike out everything you control, paid for or pitched into existence, then strike every passing mention; what survives is your real case. Our source reference covers each source type in detail.

How do companies, founders and creators compare?

Companies and founders are judged more sceptically than creators, and notability does not transfer between them. A famous company does not make its founder notable, and a notable founder does not make the company notable. Each needs its own independent coverage.

• Companies and startups are the high-risk lane, because available evidence is disproportionately marketing. The fix is genuine independent attention, which often means waiting; see Wikipedia for startups.
• Founders are judged as people under WP:BIO. "I built a successful company" is a business fact, not personal notability; see companies, founders and public figures.
• Creators and cultural figures have a structural advantage: independent reviews and retrospectives are exactly the in-depth coverage Wikipedia prefers, which fits the higher acceptance in the public benchmark.

Not sure which lane you are in? The company notability guide walks through it.

Why do freshness and independence move the odds?

Independence is tested first and forgiven least, and freshness here means spread over time rather than recency. A source is non-independent if the subject created, paid for or distributed it, or was interviewed without independent analysis. The most common reason a confident applicant is declined is that the "best" sources are well-written PR.

Coverage from a single launch week reads as a campaign; coverage by different independent authors across years reads as durable significance. A quick test: plot your strongest sources on a timeline, ask of each "did we cause this to exist?", and count the distinct independent authors writing about you in depth. Pages resting on one news cycle are over-represented in deletion discussions a year later, a dynamic covered in page maintenance and monitoring.

Is English Wikipedia the hardest edition?

For commercial subjects it is generally the most demanding, and notability in one edition does not transfer to another. Each language edition is a separate community with its own norms. English is also the highest-value edition for reach and AI visibility, so reward and risk rise together. If a subject's strongest independent coverage is in national-language press, a different edition can be the more defensible first page; see multilingual strategy. We do not quote per-edition acceptance rates, because we have not found a clean public dataset for them.

What changed with AI-written drafts?

Wikipedia is rejecting them. The Articles for Creation guidance says drafts generated entirely by large language models will be rejected, and reviewers increasingly flag the patterns: invented citations, generic filler and sources that do not say what the draft claims. Such drafts fail twice, on that criterion and on verifiability, because even a true claim must have been published in a reliable source. Polished text on a thin source base attracts more scrutiny, not less. The fix is human source research and neutral human drafting; see the 2026 AI content ban. We do not submit AI-generated text to Wikipedia.

How do you check your own risk in five minutes?

List, strike, count. This is our own rule of thumb, not a scientific model, and it assumes honesty.

1. Inventory. List every source that mentions your subject: link, outlet, date, one line on what it says.
2. Strike the non-independent. Press releases, sponsored and contributor posts, paid placements, your own channels, interviews with no independent analysis.
3. Strike the trivial. Passing mentions, directory entries, funding-round line items.
4. Count what survives. Zero or one strong independent source: not yet, build coverage first. Two or three: fixable, and an audit can say whether it is enough. Four or more in-depth ones spread over time: a real project is plausible.
5. Adjust for topic and timeline. Demand more if you are a company or startup. If all sources come from one week, treat the result as one band weaker.

Landing on "not yet" is the checklist doing its job: it saves a public decline and a doomed attempt.

How WikiBusines helps

We assess before we sell, and we say "not yet" when the sources say so.

• Notability audit, from €490, credited toward the page if ordered within one month: your sources read against policy, with a written verdict.
• Page creation, from €2,400 for English, 4 to 5 weeks. Paid editing is always disclosed on Wikipedia in WikiBusines' name, with 90 days of monitoring and guarantee after publication; if a published page cannot be defended after 3 attempts in that window, 80% is refunded.
• Annual support, from €720 per year: up to four updates, with a 12-month guarantee.

We never guarantee that an article will be published or what an AI will say about it. Start with the free 10-minute assessment: contact us. If the case is unclear, we offer the paid notability audit from €490, credited toward the page if you order it within one month.