Most AEO advice, including a lot of what we've written, treats "getting cited" as the finish line. It isn't. We've watched AI engines name a brand accurately and still do real damage in the same breath, because the source material behind that citation carried a negative signal the AI faithfully reflected.

A citation isn't automatically good. It's only as good as what it's citing.

How this actually shows up

The pattern is consistent enough to describe generally, without naming specific brands: a real Reddit thread exists where someone asks about a product, an actual user replies with a specific complaint, and the company never responds. Months later, someone asks an AI engine the same category question. The engine names the brand, because it's a real, known entity, and pairs the mention with something close to "though some users report issues with X" or "an alternative worth considering if you've had problems with Y."

That's not the AI being unfair. It's reflecting exactly what's actually been said publicly, unanswered, for however long the thread has been sitting there. The citation happened. The outcome was still a net negative.

Why silence reads worse over time

A single unanswered complaint in a live thread is a small, recoverable thing. The problem is what happens when it's never addressed: it becomes the most complete, most citable account of that specific issue. No competing description exists to balance it, because nobody from the company ever added one.

This is the same underlying mechanism behind why an unanswered thread costs more than it looks like it should. A citation engine doesn't invent sentiment. It reflects whatever's actually sitting in the source material it draws from, and an uncorrected complaint sits there indefinitely.

What actually fixes it

Not deletion, and not arguing. Neither is available to you on someone else's Reddit thread, and both would read badly even if they were.

A direct, specific, non-defensive reply to the actual complaint. Acknowledging the issue, explaining what changed or what the real tradeoff is, and being honest about anything that's still a genuine limitation. This gives future readers, human and AI both, a second, more complete data point to weigh against the first.

Timing matters more than most teams assume. A reply added six months after the original complaint still helps, but a reply added while the thread is still active reaches more of the people who'll ever read it, and reduces the window where the complaint sits as the only account.

Consistency across the record. If the same specific complaint appears in multiple places, a review site, a forum, a Reddit thread, addressing it in one place while ignoring it elsewhere leaves the pattern only partially corrected.

What this means for how you prioritize

Not every unanswered thread is equally urgent. A neutral "what should I use" thread with no company mentioned yet is a visibility opportunity. A thread where your brand is already named alongside a specific, credible complaint is a sentiment risk, and it's worth treating as higher priority, since the cost of continued silence compounds specifically in that case.

This is part of why checking actual comment threads matters, not just search snippets of the original post. The comments are usually where the sentiment actually lives.

Getting cited is necessary but not sufficient. What the citation says about you is the part that actually determines whether it helps or hurts, and that's set by what's already public, answered or not, at the moment an AI engine goes looking. We check for this specifically when we run a visibility audit, not just whether you're mentioned, but what you're mentioned alongside.