Check this page with an assistantOpens a chat asking it to summarise this article and name the evidence behind each claim.
Claude opens with the prompt on your clipboard: Anthropic does not support prefilled prompts on the web, and we would rather copy it than ship a button that drops it.
Why This Tactic Is Being Sold Again
Research keeps finding that assistants cite third-party sources far more than brand-owned pages, and that community discussion is heavily represented among them. The obvious inference for a certain kind of vendor is to manufacture the community discussion, and services offering exactly that have multiplied.
The inference is not stupid, which is what makes it dangerous. If being mentioned in a thread influences whether you get recommended, then buying mentions looks like buying influence. It is the same reasoning that produced paid link networks, and it fails for a related reason: the value of the source comes from its independence, and purchasing it removes the property being purchased.
What makes this version worse than paid links is where it fails. A discounted link costs you the fee. A discovered astroturf campaign produces a public record on the platform where your buyers are, and that record is the durable asset rather than the mentions.
The Asymmetry, Stated Plainly
The upside is small and temporary. The downside is public, permanent, and ranks. Almost no other tactic in this field has a payoff structure this bad, and it is the reason to refuse it regardless of ethics.
| If it works | If it is caught |
|---|---|
| A few mentions in threads, until removed | A thread documenting the attempt, which ranks for your name |
| A small, unmeasurable citation effect | That thread becomes a cited source about your company |
| Nothing you can report honestly | Bans across the communities where your buyers are |
| No durable asset | Regulatory exposure for undisclosed endorsement |
| Deniability, briefly | A permanent search result you cannot remove |
The second row is the one people miss. Assistants disproportionately cite community discussion, which means a well-upvoted thread about your company faking recommendations is exactly the kind of source that gets retrieved and summarised when someone asks about you. The tactic aimed at influencing citations ends up supplying a citation.
How It Gets Detected
Not by catching a single post. By noticing a pattern, which is considerably easier than the people running these campaigns assume, and which volunteer moderators have years of practice at.
The recurring fingerprints are consistent: accounts created recently with no history before recommending a product, similar sentence structure across supposedly unrelated users, activity clustered into working hours in one timezone, accounts whose entire contribution history touches one company or one category, and enthusiasm that arrives in bursts rather than distributed the way genuine mentions are.
Communities also compare notes. A pattern noticed in one subreddit gets posted in another, and the resulting thread is usually more engaging than anything the campaign produced. That is the mechanism by which a modest marketing budget converts into a durable reputational result, and it is well documented enough that anyone selling this service knows about it.
The Legitimate Version of the Same Goal
The underlying objective, being genuinely present where your category is discussed, is sound. What has to change is that participation is real, disclosed, and useful, which is slower and does not scale.
A named person from your company answering questions honestly, including recommending competitors where they fit better, builds the standing that manufactured enthusiasm imitates. It compounds rather than degrading, it survives scrutiny, and it produces the same kind of mention the campaign was trying to fake. The constraints and how communities respond are in Reddit without getting banned.
On reviews specifically, ask consistently and never incentivise a positive outcome. Asking every customer for an honest review is permitted and effective; asking for good ones, or offering something in exchange, breaches most platform policies and is detectable in aggregate. The mechanics are in how to get more Google reviews.
If Someone Is Doing It To You
Astroturfing runs in both directions, and fake negative campaigns are among the few reputation attacks that genuinely work. The response is documentation and reporting rather than public argument.
Collect the pattern rather than individual examples: account creation dates, phrasing similarities, timing clusters, and contribution histories. Platforms act on coordinated inauthentic behaviour and generally want the evidence, and review platforms have reporting routes for exactly this. That is a considerably better path than responding to each post.
Resist accusing a competitor publicly without conclusive proof. The thread that follows is about the accusation, it ranks, and it is a poor result even when you were right. This sits alongside hacking and scraping in the small set of attacks that actually damage sites, as covered in negative SEO.
Questions About Astroturfing
- What is astroturfing in marketing?
Manufacturing what looks like independent enthusiasm: paid or fabricated recommendations posted as though they came from ordinary users. The name comes from fake grass, because it is synthetic grassroots. It covers bought reviews, incentivised community posts, and agencies posting as customers.
- Why is astroturfing worse than other tactics that fail?
Because the failure is public and permanent. A bought link that gets discounted costs you the money. A discovered astroturf campaign produces a thread documenting it, and that thread ranks for your brand name, gets cited by assistants, and outlives the campaign by years. The downside is not wasted spend, it is a durable negative result.
- How do communities detect it?
Patterns rather than individual posts. New accounts with no history recommending the same product, similar phrasing across multiple users, posting schedules that look like work shifts, and accounts whose only contributions mention one company. Communities have seen every variation and moderators are unusually good at this.
- Is it illegal?
Undisclosed paid endorsement breaches advertising rules in most markets, and regulators have taken action over fake reviews specifically. Beyond regulation it breaches the terms of essentially every platform where it happens. This is one of the few tactics in this field with consequences outside search.
- What should I do if a competitor is astroturfing?
Report it to the platform with the pattern evidence rather than engaging publicly. Platforms act on coordinated inauthentic behaviour, and moderators generally want to know. Publicly accusing a competitor without conclusive evidence tends to produce a thread about the accusation, which is a poor outcome even when you are right.
