How to Spot Bought Engagement on Crypto Twitter

The first time I got burned by this, I didn't lose money on a trade. I lost time — a few weeks paying attention to an account because the replies under every post looked like a small, engaged community, and it turned out to be a rotating cast of the same forty or so accounts saying some version of "great alpha 🔥" under everything, including posts that had no alpha in them at all. Once you notice that pattern you can't stop seeing it, and once you can't stop seeing it, a follower count stops meaning anything to you on its own.

Why this matters more in crypto than almost anywhere else

On most of X, inflated engagement just makes someone look more interesting than they are. In crypto, it does something worse: it launders credibility onto whatever the account posts next. An account with a loud, apparently-real audience can move a small-cap token's price just by mentioning it, and a chunk of the people buying in are reacting to the crowd underneath the post, not to the post itself. If that crowd is rented, the signal everyone's reading isn't a signal. It's a prop. That's the specific reason it's worth learning to spot, and it's why this is one of the first things I check before I decide an account is worth following, let alone acting on.

The follower count tells you almost nothing

Start by ignoring it. Follower count is the easiest number to inflate and the hardest one to verify without tools most people don't have. It's also the number an account controls the most directly — buy followers once, and it just sits there as a permanent, static credential nobody checks the age of. If you're evaluating an account, the follower count should be the last thing you look at, not the first, because it's the one number that tells you the least about whether real people are actually reading.

What actually gives it away

The reply-to-like ratio is flat across everything

Real engagement is lumpy. A genuinely interesting post gets ten times the replies of a boring one from the same account, because real people only bother typing something when they have something to say. Bought engagement doesn't know the difference between a good post and a filler post — it reacts to both at close to the same rate, because it's paid to react, not to read. When I see an account where every post, regardless of subject or quality, pulls a suspiciously similar reply count, that's the first flag.

The replies could be pasted under any post

Read the actual words. Real replies reference something specific in the post — they disagree with a number, ask about a detail, argue with the take. Bought replies are generic enough to work under literally anything: "this is huge," "needed this today," "underrated take." One or two of those under a post is normal; X is full of low-effort real replies too. What's not normal is a wall of them with nothing specific in any single one, especially when the post itself contained something specific to react to and nobody did.

The same names keep showing up, and their own timelines are thin

This is the check that takes the most effort and pays off the most. Click into five or six of the accounts replying. A real, if small, community has people with actual timelines — their own posts, their own opinions, activity outside the one account you're checking. A rented crowd has accounts that mostly retweet and reply to a narrow rotation of the same handful of larger accounts, with little of their own. You'll start recognizing the same usernames across completely unrelated accounts' reply sections, which is close to a smoking gun — a normal person doesn't organically become a top replier on ten unrelated crypto accounts in the same week.

Engagement and impressions move in different directions

When impression counts are visible, this one's fast. Bought likes and replies are often bought separately from bought views, by different services with different pricing, so you sometimes get a post with real-looking view counts and an oddly small amount of actual engagement relative to how big the account claims to be — or the reverse, huge engagement on a post that clearly didn't reach very many people. A healthy account has those two numbers move together in a roughly consistent ratio across its recent posts. When the ratio jumps around wildly from post to post with no obvious reason, something's being juiced independently of the rest.

What this isn't about

None of this is a case for cynicism about every account with a big following. Most large crypto accounts earned it the boring way, by being consistently right or consistently useful for long enough that a real audience accumulated. The point isn't to assume everyone's faking it. It's to stop treating the raw numbers as proof of anything, because the numbers are the easiest part of an account to fake and the part most people never check before they act on what they're reading.

The numbers are the easiest part of an account to fake, and the part almost nobody checks before acting on it.

What I check instead

I've mostly stopped trying to evaluate an account from the outside in one sitting. What I do instead is watch how it holds up over time — does the same handful of people keep showing up in the replies with something real to say, do the calls or claims age well when I check back on them later, does the account ever admit being wrong about something. That takes longer than reading a bio and a follower count, but it's the only version of due diligence that isn't trivially gamed. It's also, not coincidentally, close to the definition of an actual relationship rather than a number — which is the whole reason I keep a running list of the specific accounts that have held up over months, instead of relying on a feed to remind me who was worth listening to.