There is a wonderful feature built into much of the ad-fraud industry:
Nobody really knows exactly why something was blocked.
That sounds like a flaw.
It can also be a business model.
Traffic gets labeled "invalid."
A click gets rejected.
A publisher gets told there was "suspicious activity."
Someone mentions bots, proxies, automation or anomalous behavior.
And then the conversation usually ends there.
Very scientific.
The problem is that ambiguity disproportionately benefits the party making the determination.
If a fraud vendor tells you a click is invalid, what exactly are you supposed to do with that information?
Ask why?
You may get a reason code.
"Proxy."
Excellent.
Which signal caused that determination?
How strong was the signal?
Was it the IP?
Browser behavior?
Historical pattern?
Network characteristics?
A combination of seventeen things?
Was the user definitely invalid, probably invalid, somewhat suspicious, or did the algorithm simply wake up grumpy that morning?
Usually, you do not know.
And when you do not know, you cannot meaningfully challenge the classification.
That is enormously convenient.
Because once a system becomes a black box, the output begins to acquire an almost religious quality.
The machine has spoken.
Do not question the machine.
The machine has an ISO certification.
This matters because fraud detection is not operating in an academic vacuum.
There are financial consequences attached to these decisions.
Advertisers want lower fraud.
Platforms want to demonstrate cleaner traffic.
Networks want to protect relationships.
Fraud vendors want to prove their systems are effective.
Everybody has an incentive to show that fraud is being identified and removed.
Very few people receive a bonus for discovering that some of the traffic previously condemned as fraudulent was actually legitimate.
Therein lies the problem.
Ambiguity allows every participant to tell a comfortable story.
The advertiser says fraud was prevented.
The platform says quality improved.
The fraud vendor says its technology worked.
The publisher gets less money.
Case closed.
Except perhaps the classification was wrong.
And without transparency, nobody can tell.
This is why the industry needs to move beyond unexplained labels.
Note what the problem actually is. It is not that the answer is binary.
A click either came from a browser lying about what it is, or it didn't. That is not a matter of degree. That is a fact, and facts are allowed to be binary.
The problem is that nobody shows you the fact.
"Invalid" with nothing behind it is useless.
"Invalid — this browser reported itself as Safari on iOS, then answered a question only Chromium can answer" is something you can act on. You can take it to the partner who sold you the traffic. You can watch them try to explain it.
Adding a confidence score does not fix this. It makes it worse.
Because "invalid, 73% confidence" is not more transparent than "invalid." It is less. You have now been handed a number you cannot audit, cannot challenge and cannot take anywhere, and the vendor has acquired a permanent defence: you set the threshold.
A probability is not evidence.
A probability is a way of not saying what you found.
So the demand should not be "give me a confidence level."
The demand should be:
What did you observe, and why does it mean what you say it means?
What signals were observed?
What specifically contradicted what?
Can I check the reasoning myself?
Can I take it to the counterparty and have the argument?
Because ambiguity is not neutral.
When one party controls the definition, the evidence, the classification and the explanation, uncertainty tends to flow in only one direction.
And strangely enough, so does the money.