Click fraud case study: Adwords and YSM compared

Click fraud trackingI’m keeping this post in english since some english speaking people might be interested as well.

Recently I have opened a case with Yahoo Search Marketing concerning click fraud. Though I sent them all the data they needed they came back to me with some standard questions which I covered already in the first mail. Additionally they stated they can only investigate on date ranges of 60 days. Keep in mind I sent them detailed data of 6 months! Uhm.. ok. I assume they want me to open three cases.
When replying to YSM I did some further investigations which I’ll show below:

  • Campaign: same landing page for both PPC engines
  • Date range: 1/1/09 – 1/16/09

Adwords
Clicks received: 100%
Fraudulent/duplicate clicks according to tracking system: 35%
Reported (and payed!) clicks according to Adwords: 73%
Discrepancy: 8%

YSM
Clicks received: 100%
Fraudulent/duplicate clicks according to tracking system: 35%
Reported (and payed!) clicks according to YSM: 110%
Discrepancy: 45%

It’s interesting to see the same percentage of click fraud with both PPC engines. But it seems Google does a better job in filtering these clicks. Taking into account that YSM sends only a fraction of the traffic Adwords does Yahoo Search Marketing is almost not worth the time spent for setting up and managing campaigns. For this specific campaign CPC in YSM is even way higher than CPC in Adwords.
Alltogether this results in a ROI of 111% with Adwords and 57% with YSM. Not to mention YSM is responsible for only approx. 10% of the net profit while the rest is ascribable to Google.

I really think it’s important there’s some competitor for Google Adwords and one should support the smaller PPC engines. But if they think they can send less traffic and charge more for it I have to re-think my attitude!

What is your experience on this subject?

3 Kommentare zu „Click fraud case study: Adwords and YSM compared“

  1. Hey Eric, what’s up?

    That’s a really great article, congratulations.

    By the way, I can’t recognize that report, where is it from?

    Keep on the good work.

    Joseph.

  2. Eric Marchionni

    Yeah, that’s what I thought too when I had made the math.

    The calculation is based on the filtered out clicks of prosper202. I’m aware this only hits the obvious duplicate/fraudulent clicks. More intelligent click fraud software which makes use of proxies won’t get caught.

    Whats your experience with click fraud?
    What’s the fraudulent click rate if you do track it?

Kommentar verfassen

Deine E-Mail-Adresse wird nicht veröffentlicht.