
HonestHook
Exact social data from 9 platforms, one API
You ship it, it returns numbers, you move on. Then Threads rotates its TLS fingerprint. Bluesky starts refusing your IP range. A field you've been reading for six months quietly starts pointing somewhere else, and nothing errors - the number just goes wrong.
Writing the scraper was never the cost. The babysitting is. And the second problem doesn't announce itself either: any history you didn't collect yesterday doesn't exist. You can't backfill last Tuesday's follower count.
I don't write code. I direct an AI to write it, and the one rule I gave it was that the number has to be the account's real number, not the first one that pattern-matches.
Here is what that means in practice. Measured on pinterest.com/nasa on 2026-09-08: the string follower_count appears 42 times in a single profile page, with different values. 41 of them belong to boards the account owns, and a board's count looks exactly as plausible as the account's. So the Pinterest route doesn't pattern-match at all - it reads the ProfilePage structured-data block the page publishes.
Same shape of problem on TikTok: stats.followerCount is the first follower field in the payload. statsV2.followerCount is the one the profile page actually shows. A first-match read hands you the wrong one.
A real response, unedited:
200 OK - 1 credit
{
"success": true,
"platform": "tiktok",
"endpoint": "tiktok/profile",
"data": {
"author": {
"handle": "nasa",
"followers": 1692680,
"following": 23,
"posts_count": 44,
"verified": true
}
}
}
Nine platforms behind one schema: Instagram, TikTok, Threads, Bluesky, Pinterest, SoundCloud, Mastodon, GitHub, Linktree.
Free tier is 1,000 credits a month, no card. One profile call is one credit, so that's 1,000 profiles a month before you pay anything. A call that fails isn't charged.
Straight with you: zero paying customers so far. The name is a promise about the data, so I'm not going to open my first post here by implying traction I don't have.
Live demo you can run against your own handle: https://honesthook.com/demo?p=ih
For anyone who has kept a scraper in production - what broke it last for you: the platform changing shape, or your own backfill gap?
About
Every social API rounds the numbers or hides them, and running your own scrapers is a full-time job. I wanted one API that returns exact public data across 9 platforms, so I built it. No paying customers yet.

1 Comment
The technical evidence makes the reliability problem credible, but zero paying customers leaves the commercial question open. Have any developers actually replaced a scraper they were maintaining themselves with HonestHook, or is usage still mostly testing the data quality?