Esuba vs fraud scoring.
Fraud scoring and Esuba work on opposite ends of the same problem. Scoring judges each signup as it arrives. Esuba goes after the toolkit producing the signups and takes it down. Most AI and API companies being farmed need both.
At a glance
| Fraud scoring | Esuba | |
|---|---|---|
| Sees | Each signup, as it happens | The toolkit, the operator and their infrastructure |
| Acts by | Blocking or challenging the signup | Reporting the toolkit to the platforms and hosts that run it |
| Against a farm | Raises the cost of each account | Takes the toolkit away and tracks its successors |
| Other targets | Not covered | Warns every company the same operator is hitting |
| Your data | Needs every signup | Public sources; your data is optional |
| Your code | An integration | Untouched |
| Price | Set by each vendor | $2,000 a month, everything included |
Where scoring is strong
Scoring stops a lot of low-effort abuse before it costs you anything, on every signup, with nobody watching. If you don't have it, you probably should.
Where it runs out
Scoring sees the signup, never the toolkit behind it. A farm built to beat your checks adapts: in Case 001 the operator moved between three signup routes in five days, and each move reset the cat-and-mouse.
Scoring also can't touch the farm outside your walls: the repos, the domains, the channel selling access. Those keep working until someone reports them.
Using both
Keep your scoring. Esuba finds where the farm is built, reports it, and tells your team which patterns to feed back into your rules.
Questions
Do we need Esuba if our fraud scoring works?
If no toolkit is aimed at you, maybe not. Our free farm exposure report will tell you whether one is.