Paste a tweet. See the volume it brought in, whether the engagement was real, and how it compares to every call we've measured.
Free — no account for your first checks
Across everything we've measured, the audience that actually trades has almost nothing to do with audience size.
A 15x gap in volume brought per follower — completely invisible if you're buying on reach.
Know which posts moved money before you pay for the next round.
Prove your numbers instead of quoting a follower count.
See whether the account you follow actually calls anything.
Median impact across every scored call is 8.6 out of 40. Most KOL posts do nothing.
A 28.8K-follower account brought $3,018 per thousand followers. A 165.6K account brought $191.
Audience quality does not predict returns. We tested it and the correlation is 0.05.
KOL skill does not persist. First half against second half of the same account's history: −0.02.
Reply-content scoring beat account-stat scoring at 91% against hand-labelled posts.
Roughly 31% of a KOL's token posts are not calls at all.
Audience quality has no measurable relationship with returns — we tested it, and there isn't one. Blending them into a single score would imply a link the data denies, so we keep them apart.
Volume and distinct traders brought in against the pair's own baseline, with attribution decaying by how soon the move arrived. A peak fourteen hours later is not the post's doing.
Peak, hold and drawdown measured in excess of peer tokens on the same chain — so a market-wide melt-up never reads as skill.
Reply content over account stats: wallet-address giveaway entries, contentless praise, absence of substance. Validated at 91% against hand-labelled posts.
Engagement and volume, both per thousand followers, against the measured p75 of each. Separates the quiet audience that trades from the big one that only makes noise.
Charts rotate through real posts from the corpus. Hover for the underlying numbers.