Wikipedia grades every article on a quality scale that has nothing to do with why the article exists in the first place. Stub, Start, C, B, Good Article, Featured Article — a small badge sits on the article's talk page recording where it falls, assigned by whichever WikiProject has adopted the topic. The scale measures prose, structure, citation density, image coverage. It says nothing about notability, and it isn't meant to.
A regional engineering firm's article ran four sentences: founding year, headquarters, principal service line, one paragraph on a bridge project covered by two trade publications. Stub-class, by any WikiProject's rubric. The firm's marketing director read the grade as an insult and asked us to expand it — add the leadership team, the client list, the mission statement, three more paragraphs of history sourced to the company's own annual reports.
None of that would have moved the article out of stub class in any way that mattered, and most of it would have made the article worse. Wikipedia's notability guideline asks a single question: has an independent source given this subject substantial attention. A mission statement sourced to the company's own website answers a different question, and a reviewer who finds three paragraphs like that sitting next to two genuine trade-press citations tends to draft a maintenance tag before they draft a compliment.
Padding a short article exposes it to a different risk than the one padding is meant to solve. Both the notability guideline for organizations and the general one ask about the sourcing that exists, not the sourcing that's convenient. Filling a four-sentence article with three unsourced or self-sourced paragraphs takes a stub that would sail through a deletion discussion and turns it into a mixed record a reviewer has to sort through paragraph by paragraph.
Article length should follow the sources that exist, not the ambition of the person writing it.
What moves a quality rating is genuine content: additional independently sourced facts, a lead that accurately summarizes a longer body, better-structured sections. What decides a deletion outcome is something else, narrower: whether the sources underneath any of it are independent, reliable, and substantial. A four-sentence stub built on two solid trade-press articles clears that bar. A two-thousand-word entry built on the company's own materials, a sponsored interview, and one passing mention does not, regardless of its class rating or its length.
We tell clients the quality badge is a maintenance label, not a verdict, and that the sentence count is the wrong thing to negotiate over. The question worth asking before adding anything is whether it comes from a source with no reason to inflate the subject, not whether the article currently reads as thin. Thin and accurate outlasts long and undocumented, in almost every discussion we have watched.
The engineering firm's article stayed at four sentences for another eight months, until a regional business paper ran a longer piece on the same bridge project. That paragraph went in, sourced to the new piece, and the class badge moved to Start. Nothing else about the article changed. The subject hadn't grown. The record had.
