Bot Farms: How Social Signals Are Faked and How to Detect Them

Fake social activity is a staple expense of any pump campaign: simulating "organic hype" costs pennies and is purchased as a service. Being able to spot it is basic literacy.

What pumpers buy

• Reply waves: hundreds of comments with the ticker under viral tweets - the main channel for the crowd to "discover" a token.
• Hashtag trends: synchronized posting pushes the ticker into trends - platform algorithms then work for free.
• Channel metric padding: views and "reactions" on TG, upvotes on Reddit - to build a showcase for call buyers.
• Armies of "enthusiastic holders": accounts posing as a living community in the token's chat.

Detection in seconds

Profiles: fresh accounts, generated nicknames, zero personal life, mass subscriptions to the same tokens. Text: templates with the ticker plugged in, identical emoji patterns, calque translations. Timing: a wave of activity within a single minute - real people don't behave that way. Layer mismatch: thousands of "discussions" against a hundred on-chain holders - the numbers lie somewhere.

Our social radar weighs mentions by account quality (age, followers, history) for exactly this reason: a raw mention counter is a metric for victims. Verified signals are in the section.