There's actually quite a buzz in the online world about this "audit" from the Chitika folk prior to their November payments, actually, so you're definitely not alone.
The numbers are so high, however, that it's extraordinarily odd. With some of my colleagues, it's really hard to imagine that they have even a single fraudulent click, let alone 30-50% being classified as fraudulent by the Chitika folk.
I dug around a bit to see if I could ascertain what algorithms they use to determine whether something is fraudulent or not, and found this interesting passage in their terms of service:
"Chitika shall not be liable for any payment based on (a) any fraudulent impressions generated by any person, bot, automated program or similar device or for fraudulent clicks similarly generated on any Paid Listing, as reasonably determined by Chitika and or its Partners; (b) Paid Listings delivered to end users whose browsers have JavaScript disabled; (c) Paid Listings benefiting charitable organizations and other placeholder or transparent Paid Listings that Chitika may deliver in the event that a Site is improperly configured to comply with Chitika technical requirements; (d) Chitika or its Partner advertisements for its own products and/or services; or (e) impressions co-mingled with a significant number of fraudulent impressions or fraudulent clicks described in (a) above, or as a result of other breach of this Agreement by Client for any applicable pay period."
Interesting, but here's an even more interesting snippet (with my highlights):
"Chitika reserves the right to withhold payment of any Revenue Share Payment in the event of any breach of this Agreement by Client, pending Chitika's reasonable investigation of any of the foregoing or any breach of this Agreement by Client, or in the event that an advertiser whose Paid Listing are displayed on the Site defaults on payment for such Paid Listing to Chitika."
I have also heard through the grapevine that Chitika automatically rejects any clicks from Asian countries as part of its sweeping anti-fraud strategy, so if you did happen to attract a large audience from, say, Japan, that could be a big problem.
The other issue is "impressions co-mingled with a significant number of fraudulent impressions or fraudulent clicks". Maybe it's just me, but doesn't that suggest that if their fraud algorithm detects fraudulent clicks for whatever reason that all clicks during that period of time are discounted, even legitimate traffic?
In any case, what baffles me is why Chitika doesn't just filter its traffic numbers on the daily reports against the "typical" fraud numbers so that if the average Chitika fraud rate was, say, 25%, then they should just automatically subtract 25% from the numbers they display. Then if you had a typical fraud rate, you'd have accurate numbers, and if you had less than the typical fraud rate (one pal of mine tells me he was penalized only 13%) you'd have an end-of-month increase, which would be darn nice.
Anyway, perhaps someone from Chitika can pop on here and give us a bit more of a definitive answer on how they calculate fraud rates? We'll hope for the best.
Hope this helps you understand what's going on with your Chitika account.
Sign up for my weekly AskDaveTaylor Newsletter and you'll receive even more tech and gadget help
right to your inbox, along with exclusive news and industry updates. It's good stuff. I promise!
I think Dave hit the nail on the head in that last point, which is: why doesn't Chitika just adjust your "revenue" at the time that the "unaudited reports" are generated? It could be more accurate than just subtracting 25% from your expected revenue; Chitika could just subtract out the clicks that they know are going to be disqualified later (clicks from Asia, or clicks from duplicate IP addresses).