Wikipedia rejections at Articles for Creation follow predictable patterns and most of them trace back to issues that should have been caught in the readiness assessment. Insufficient notability is the most common: the sources do not support the article’s claim to encyclopedic importance, usually because they are too few, not in-depth enough, or not independent of the subject. Promotional tone is next: language that reads as marketing copy, undue emphasis on awards or rankings, subjective adjectives describing the subject. Undisclosed COI gets the article flagged and often declined regardless of other merits. And insufficient secondary independent coverage means the third-party authoritative material is too thin to support the substance of the article. Each is fixable, but the fix is usually upstream of the article itself, in the sourcing and disclosure work that should precede submission.

Wikidata is a free, structured knowledge database maintained by the Wikimedia Foundation and run by the same broader community as Wikipedia. For reputation work, Wikidata matters for two compounding reasons. First, it is one of the primary inputs to Google’s Knowledge Graph, so improvements to the Wikidata entry flow into the Knowledge Panel directly. Second, AI engines use Wikidata for entity disambiguation, so a clean Wikidata entry with proper identifiers and sameAs links improves how the engines recognize and describe the entity. Wikipedia and Wikidata are maintained together; getting one right without the other leaves value on the table.

The Wikipedia sandbox is a drafting workspace, located at User:[username]/sandbox for any registered editor, where article content can be developed without being live in mainspace. The sandbox is where the drafting work actually happens for a new article: the article structure gets built, sources are added inline, the language is tightened to encyclopedic voice, and the content is reviewed before any external eyes see it. From the sandbox, the article can be moved to mainspace directly (faster but bypasses AfC review) or submitted to Articles for Creation for community review before publication. For COI work we work in the sandbox first, refine through internal review, and then submit through AfC with the COI disclosed. The sandbox does not protect against later challenges to the article but it does protect against the workflow problem of editing live and then trying to fix issues in public.

Stubs and full articles are different stages of the same article rather than different types. A stub is a short article, typically under 500 words, that covers the basic facts about the subject – what it is, when it was founded or born, where it is located, what it does or did – and is sourced enough to establish notability but not enough to support a full treatment. Stubs are tagged as stubs and explicitly marked as incomplete, with a maintenance template inviting expansion. A full article is comprehensive: developed sections covering the subject’s history, structure, key activities, and relevant context; multiple sourced sections; conformance to the Wikipedia Manual of Style; and breadth of coverage that matches the substance of the subject. For COI work, starting with a stub is sometimes the right call when the sourcing supports notability but not full development; the article gets created, the entity gets the Knowledge Panel benefit, and the expansion happens over time as more sources accumulate.