Three terms, defined once. Notability is Wikipedia's test of whether a topic deserves its own article. AfC (Articles for Creation) is the review queue where drafts are accepted or declined. AfD (Articles for Deletion) is the public discussion where a published article can be kept or removed. COI is conflict of interest: editing about yourself, your employer or your client.
How do the three deletion routes differ?
"Deletion" is not one process but three, and knowing which you face changes the response. The fastest are the most dangerous.
| Route | Trigger | Speed | Who decides | Policy |
|---|---|---|---|---|
| Speedy deletion (CSD) | Blatant promotion, copyright violation, no credible claim of significance | Hours, "without waiting for any discussion" | One administrator | WP:CSD |
| Proposed deletion (PROD) | Uncontroversial removal | 7 days | An admin, only if nobody objects | WP:PROD |
| Articles for Deletion (AfD) | Contested notability or sourcing | "At least seven days" of discussion | Community consensus, closed by an admin | WP:AFD |
PROD is the gentlest: any editor may cancel it by removing the tag. AfD is an argument you can win on the merits if the sources exist. Speedy deletion is the one to fear, because a promotional draft can vanish before anyone on your side sees the tag. The general principle behind all three is the deletion policy: pages that do not meet the criteria for encyclopedic content are removed.
Is a rejected draft a writing problem or an evidence problem?
Usually evidence. Two drafts can look identical on the page, thin and promotional, yet need opposite cures: one needs an editor, the other needs a year of earned press before it should be attempted at all.
We classify every failure on two axes. The layer says where it broke: the evidence layer (sources cannot support an article), the representation layer (sources exist but the draft mishandles them) or the process layer (the way the article entered Wikipedia was flawed). The fixability says whether a competent, disclosed editor can solve it with materials you already have (process-fixable) or only new independent coverage can (evidence-fixable). This is our working model, not official Wikipedia terminology.
What are the 12 failure patterns?
Each pattern below names its layer, the governing policy and its fixability. Read fixability first: it tells you whether to call a writer or your PR team.
1. Weak notability
Evidence. WP:GNG and WP:NCORP. Evidence-fixable.
The largest single cause. A topic is presumed suitable for an article when it has received "significant coverage in reliable sources that are independent of the subject". For organizations the guideline states that "no company or organization is considered inherently notable". Revenue, headcount and a good product are not notability; coverage about you by people unconnected to you is. If it does not exist, rewriting changes nothing.
2. Promotional tone
Representation. WP:NPOV. Process-fixable.
Neutrality is described in the policy as non-negotiable. Words such as leading, innovative, award-winning assert judgments the sources have not made. Promotional drafts are declined at AfC, and blatantly promotional articles are speedy-deletion candidates. If the coverage is genuinely there, a neutral rewrite usually passes. The mistake is treating tone as the only problem when notability is also weak: you then get a tidy article that still falls to pattern 1.
3. Primary-source dependency
Evidence. WP:RS and WP:NOR. Evidence-fixable.
A reference list built from the company website, press releases, a founder's LinkedIn, sponsored content and self-serving interviews is the classic tell (more in our guide to reliable sources). Articles should rest on reliable, independent, published sources; press releases and interviews fail independence even on a respectable outlet's site, because the words originate with the subject. Wikipedia also forbids original research, so you cannot narrate your own story from internal knowledge.
4. Paid-editing disclosure problems
Process. WP:PAID and the Wikimedia Terms of Use. Process-fixable, and mandatory.
This is a rules violation however good the article is. Anyone paid to contribute must disclose it, including each employer, client and affiliation. Undisclosed paid editing can get the account blocked and the article deleted on process grounds alone, which is why a "confidential, nobody will know" pitch is a red flag. The compliant route is also the safer one: disclose on the user page or talk page, and because conflict-of-interest editors are strongly discouraged from editing affected articles directly (WP:COI), propose changes on the talk page or through AfC. See paid editing, COI and disclosure.
5. AI-generated text
Process. WP:AFC and WP:NOR. Process-fixable, with an evidence trap.
Reviewers spot machine-written drafts quickly: invented citations, sources that do not say what the text claims, generic filler. The AfC guidance says that articles generated entirely by LLMs will be rejected. The worst damage is the fabricated reference, which poisons the whole draft. The trap: AI prose can make a subject look sourced when it is not, hiding a pattern-1 failure underneath.
6. No sustained coverage
Evidence. WP:NCORP and WP:GNG. Evidence-fixable (time).
A funding round covered by five outlets in 48 hours feels like a win. Reviewers see a burst of near-simultaneous, often press-release-driven coverage that goes silent. Notability is closer to durable notice: attention that recurs, independently, over months and years. A briefly newsworthy subject should usually wait.
7. Single-event notability
Evidence. WP:GNG and WP:BIO. Evidence-fixable.
The subject is covered in depth, but only in connection with one incident: a viral campaign, a lawsuit, a controversy. Wikipedia generally prefers an article about the event over a standalone biography or company page built on it. The cure is more and more varied independent coverage showing the subject matters beyond the one event.
8. Awards and rankings that do not count
Evidence. WP:RS and WP:NCORP. Evidence-fixable.
"We won three industry awards" is common and weak, because many awards are pay-to-enter, vendor-run or "best of [city]" lists with no independent editorial selection. A selective, independently judged honor with secondary coverage of the win can help; a logo on your homepage cannot.
9. Local-only coverage
Evidence. WP:NCORP. Evidence-fixable.
Routine coverage in one local outlet, such as a business journal profiling every new shop, rarely establishes notability. For European businesses, strong national or regional press in another language does count, since non-English reliable sources are accepted, provided it is independent and substantial. For choosing the edition, see our multilingual strategy guide.
10. Founder and company conflation
Evidence. WP:BIO versus WP:NCORP. Evidence-fixable.
People and organizations are judged under different guidelines, and notability does not transfer between them. A notable company does not make its CEO notable, and a notable founder does not automatically make their startup notable. Decide which subject you are arguing for and source that subject. Our deep dive: companies, founders and public figures.
11. Poor draft structure
Representation. WP:V and WP:NPOV. Process-fixable.
The subject is notable and the sources exist, but the draft has claims without inline citations, unsourced "About us" prose and no clear lead. The burden of verifiability lies with the editor who adds the material. This is the cheapest failure to fix: same facts, same sources, correct packaging.
12. Post-publication neglect
Process. WP:DEL and WP:AFD. Process-fixable, ongoing.
A published article is not permanent. It can be nominated years later, vandalized or quietly degraded. If nobody who understands the sourcing defends it during the AfD window, it can be lost for answerable reasons, and deletion removes all versions from public view. The defense is monitoring, a source base kept current and someone ready to argue on the merits; see monitoring and the lifecycle risk curve.
The 12 patterns at a glance
| # | Pattern | Layer | Fixable by |
|---|---|---|---|
| 1 | Weak notability | Evidence | New evidence |
| 2 | Promotional tone | Representation | Process |
| 3 | Primary-source dependency | Evidence | New evidence |
| 4 | Paid-disclosure issues | Process | Process (mandatory) |
| 5 | AI-generated text | Process | Process (evidence trap) |
| 6 | No sustained coverage | Evidence | New evidence (time) |
| 7 | Single-event notability | Evidence | New evidence |
| 8 | Weak awards and rankings | Evidence | New evidence |
| 9 | Local-only coverage | Evidence | New evidence |
| 10 | Founder and company conflation | Evidence | New evidence |
| 11 | Poor draft structure | Representation | Process |
| 12 | Post-publication neglect | Process | Process (ongoing) |
Eight of twelve are evidence-fixable. That is the uncomfortable, useful result: most failures are evidence problems dressed as writing problems, which is why a source audit, not a rewrite, is the correct first spend.
How can I test my sources before drafting?
Score each source you intend to cite, 0 to 3 on four questions, in a spreadsheet, in about twenty minutes. This is our own rule of thumb, not an official Wikipedia metric.
| Question | 0 | 1 | 2 | 3 |
|---|---|---|---|---|
| Independent of the subject? | Own site, PR, interview | Mostly subject-supplied | Independent outlet, some subject input | Fully independent reporting |
| Significant coverage? | One-line mention | A paragraph | A substantial section | The subject is the article |
| Reliable outlet? | Self-published, pay-to-list | Unknown blog | Recognized trade or regional press | Major outlet with editorial standards |
| Secondary (analysis, not announcement)? | Press release, listing | Reprinted wire copy | Reported news | Analysis, profile, investigation |
- 10-12: load-bearing. You typically need three or more, about different events or angles, for a defensible company article.
- 4-9: supporting. Useful for facts, not for carrying notability.
- 0-3: decorative. If they dominate the list, expect patterns 1, 3, 8 or 9.
If you cannot find three load-bearing sources that are independent, in-depth and not clustered around one event, the draft will probably fail pattern 1, and the right move is to build coverage first. Two free self-serve tools help: the media list and the Notability Index, which shows how our own assessments (2018-2026) of notability vary by industry.
Can anyone guarantee a page will not be deleted?
No. Wikipedia is run by a volunteer community and a body of policy; no vendor controls reviewers, administrators or consensus. Even the notability guideline says significant coverage creates "an assumption, not a guarantee" that a topic merits an article. Anyone promising approval or permanence is misunderstanding the process or asking you to ignore it.
What a serious provider can do is reduce risk before money is spent: a notability assessment, real source research, neutral drafting, transparent paid disclosure under WP:PAID and post-publication monitoring. Our terms are on /guarantees: we monitor for 90 days after publication, and if a published page cannot be defended after three attempts within that window, we refund 80%. That describes our work, not control over the community.
How WikiBusines helps
If your honest read of the matrix is that you sit in the evidence column, the useful first spend is diagnosis, not drafting. Spending €2,400 to draft an article whose sources score as decorative is how money gets deleted.
- Notability audit, from €490. Your real sources read against policy, with a go or wait verdict. The fee is credited toward the page if you proceed within one month.
- Page creation, from €2,400 for English, 4-5 weeks. Neutral drafting, with paid editing disclosed on Wikipedia in WikiBusines' name, and 90 days of monitoring after publication.
- Deleted page recovery, quoted per case. For pages already removed; the plan depends on why they fell.
- Annual support, from €720 per year, with a 12-month guarantee. Near real-time monitoring and alerts come with Managed Protection at €1,500 per year.
For pattern-specific rewrites of an existing page, see editing (€700-€1,900). 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. Full prices are on /pricing, and the cost breakdown is in what a Wikipedia page costs.
Frequently asked questions
Almost always for one of the 12 patterns above, with notability failure as the dominant cause. Promotional tone, reliance on the subject's own sources and undisclosed paid editing are the next most common. The basis is the deletion policy, which removes pages that do not meet the criteria for encyclopedic content.
A single administrator may delete a page without waiting for discussion when it meets a defined criterion, such as blatant promotion or copyright violation (WP:CSD). It can happen within hours, and overtly marketing-style drafts are what trigger it.
An AfD runs at least seven days, then an uninvolved administrator closes it on community consensus (WP:AFD). Proposed deletion also runs seven days but is cancelled when any editor removes the tag.
Sometimes, depending on why it was deleted. A page removed for fixable reasons can be improved and resubmitted or taken to deletion review; a page deleted for genuine non-notability cannot return until real coverage exists. See how to get a deleted page back.
Yes. Neutrality is non-negotiable, and sufficiently promotional articles are speedy-deletion candidates. If the sources are strong, a neutral rewrite usually saves the page; if they are weak, fixing the tone alone will not.
Most often notability (pattern 1) or primary sourcing (pattern 3): the reviewer did not find enough independent, in-depth coverage. See why your draft was rejected for the decline reasons.
Show up with the sources. The article is kept or deleted on the independent, significant, reliable coverage demonstrated during the discussion, so you need a strong source base and someone who can argue it on the merits, which is what monitoring provides.
Occasionally: a notable subject packaged badly (pattern 11) is process-fixable. Treat "it's just the tone" with suspicion, though, because weak tone and weak notability often travel together.
Authors and editorial policy
Bohdan DubylovskyiCo-founder and CBO, WikiBusines→
Roman MelnykCo-founder & Chief Editor, WikiBusines→The article was prepared by the WikiBusines co-founders. We sell paid Wikipedia services, so we disclose this interest: paid editing is always declared on Wikipedia in WikiBusines' name, and claims about the rules are backed by links to primary sources.
Editorial policy: payment disclosure, conflicts of interest, sources