Last night, I began analyzing the Farcaster Pro data (basescan API + Neynar to retrieve user information using the wallet address). Found some interesting insights. I will write about a few, and attach the GitHub repos for anyone who wants to analyze them further. This is the data for the first 24h.

1195 users with 0 followers, 1289 with 1-10, and 125 with no valid acct(?). Using neynar v2/farcaster/user/bulk-by-address

Transactions plateaued after the 10K number hit. Expected, I guess.

working on aerospace applications during day time. building agents after sunset @rootnode | mani.ink

Neynar score distribution for all pro users. Median 0.5, majority in the medium range. Pro != high_quality_users I'd assume. Dunno how the metric is computed. (cc @rish )

building farcaster infra @ /neynar 🪐 | casting @ /rish

Next is follower counts. Folks with <= 10 followers are about 1/4th of the subscription. I manually went to their FC profiles (random 10 of them) to make sure it's not a bug in the code.

FID vs. score also looked interesting (note: this data is only available for Pro subscribers).

And this was for unknown users. Spiked around 1615 PST. I dont know how this works, am a crypto noob. Is it possible that a bunch of addresses just did the token transaction so that they will get the NFT airdropped? Or a bug with the API? I only tried retrieving info once. @neynar

Link for anyone who want to explore further. Make sure to keep the transaction json file in parent directory, the basescan api only returns last 10k transactions. Reply/DC if you find any bug - I tried to verify the data manually as much as I can.

Link:

Link for anyone who want to explore further. Make sure to keep the transaction json file in parent directory, the basescan api only returns last 10k transactions. Reply/DC if you find any bug - I tried to verify the data manually as much as I can.

Link: https://github.com/manimohans/fc_pro_analytics/

You can also have fun by dragging and dropping the csv file into your favorite AI. Screenshot from Gemini here.

Yes, brain naturally assumes pro == high quality content. The data showed otherwise.

anyone can buy a pro subscription, has no correlation to quality

we need some examples of who/what this actually represents I think

I also asked about this and @rish answered this

https://farcaster.xyz/rish/0x232177dc