Skip to content
UTMCAP

Click fraud detection

Click fraud detection, with the dial exposed.

Most tools give you a number and ask you to trust it. A number you cannot take apart is no use when a network disputes it, so the useful question is which signals fired and what each one is worth.

Not all bad traffic is fraud, and it helps to separate the two. Some of it is honest automation: search crawlers, uptime monitors, link previews from chat apps. It is not trying to cheat you, but you should not pay for it either. The rest is traffic dressed up to look like a person, and its whole purpose is to be paid for.

A tracker can see quite a lot at the moment of the click. Whether the address belongs to a data centre rather than a home or a phone network. Whether the browser is being driven by software. Whether something claims to be Googlebot from an address Google does not publish. Whether the same address has clicked eleven times in a minute.

What a tracker cannot see is intent. A real person paid a cent to click your ad looks exactly like a real person who wanted to. That is why the honest output is a score with reasons attached, not a verdict, and why you should be suspicious of anyone selling certainty.

How it works

The mechanism, not the marketing.

  1. 01

    Signals are checked

    Data-centre and proxy ranges, Tor exits, crawler impersonation, headless browsers, missing headers, click floods.

  2. 02

    They add up to a score

    Each signal carries a weight. The total is compared to a threshold you set rather than one we chose.

  3. 03

    The reason is kept

    Every filtered click records which signals fired, in words, so it can be shown to somebody else.

  4. 04

    It is excluded from the money

    Filtered clicks come out of cost and revenue figures, so a bot wave cannot flatter or ruin a campaign.

What to check tonight

You can do all of this without us.

None of it needs an account anywhere. If it turns out your tracking is fine, you have lost twenty minutes and learned something.

How UTMCAP does it

What you would get here.

Start free — 100,000 clicks

Twenty-one signals

Data-centre address, proxy and VPN exits, Tor, crawler impersonation, headless browsers, missing user agent, click floods and more.

Two that cannot be switched off

A request pretending to be Googlebot and a browser being driven by software are not judgement calls.

A threshold you set

Move the dial for your own traffic. Push traffic and search traffic do not deserve the same setting.

Reasons in words, per click

Which signal caught it, on the click itself — evidence rather than a total.

Allow and deny lists

Run before scoring, each entry carrying the note whoever added it wrote.

Routing on the verdict

Send suspected bots somewhere harmless instead of paying an advertiser for them.

Questions

The ones people ask next.

Read next

Do filtered clicks still count against my plan?
Yes, because deciding whether a click is a bot is the work we do. They are excluded from cost and revenue figures, so your campaign numbers stay clean.
Can I get money back from a network for bot traffic?
Sometimes, and specifics are what win it. A count is arguable; a list of clicks with timestamps, addresses and the reason each was filtered is much harder to wave away.
Is this the same as an anti-fraud product?
No. Dedicated products go further — device fingerprinting, their own reputation networks, a full fraud score. Ours is filtering that is honest about its signals, included on every plan rather than sold separately.

Point a domain at us and run a click through it.

The free plan takes a hundred thousand clicks a month and no card. If it does not do what you need, you have lost twenty minutes.