THE RKGBLOG

More Affiliate Dirty Tricks: Device Targeting Brand PPC

There’s a certain class of affiliate marketer that will go to any lengths to make a quick buck — short of providing any real value to the companies they are supposed to be supporting.  One of the tried and true tactics of this type of affiliate is to run paid search ads on a company’s brand terms, trumping the company itself by utilizing their display URL and pocketing the difference in ad costs and affiliate commission.  Generally, this behavior is expressly forbidden in the affiliate terms and conditions and monitored by the advertiser and/or its agency, so the black hat affiliate needs to get sneaky.

In the past, we have found that affiliates playing this game were only doing so in the wee hours of the night when their activities were less likely to draw attention.  Similarly, some affiliates have geo-targeted their brand campaigns to exclude not only our clients’ headquarters, but RKG’s own offices.  One of the more nefariously clever tactics we’ve seen was an affiliate spoofing RKG’s redirect in their ad destination URL, so at a quick glance it appeared to be our ad, but the affiliate had really inserted their own affiliate code into the link to take the credit and a nice commission.

The latest incarnation of this scam, for lack of a kinder interpretation, is for the affiliate to only run the brand paid search ads on specific devices like the iPad.  Again, the idea is that the advertiser, agency or monitoring service is less likely to come across the deception if it is being carried out in a niche segment.

So, how can you catch them?  Your first stop should be the good old AdWords Ad Preview Tool where you can specify the type of device you want to mimic and for which geographic location.  If the ad showing is not yours, but looks just like yours — the best practitioners of  this tactic use the exact same ad copy as you are running — it’s likely an affiliate.  In fact, affiliate versions of your ad can even show with seller ratings because they live at the domain level, rather than the account level.

Another option is to install a browser add-on that allows you to operate your desktop browser as though it were a different device.  We’re partial to the User Agent Switcher extension for Firefox.  Once you’ve installed the add-on and turned on iPad user agent spoofing, do a Google search for your trademark, right click the ad, and copy the link location.  Paste this into a text editor and see what you’ve got.  You can also do the same sort of research on an iPad itself.

In one example we uncovered, the affiliate was using a bit.ly URL that was redirecting to their affiliate site and then to the client’s site.  How handy!  Add a ‘+’ to the end of a tinyurl or bit.ly URL and you can see how much traffic the URL has received — and for how long — to get a precise view of how severe the infraction has been.

More automated means of trademark monitoring would employ similar methods of spoofing user agents and geographic locations to make sure that the rogue affiliate has difficulty hiding their activities.

Fortunately, the vast majority of affiliates operate above board.  Even if the value some of them contribute is debatable, they adhere to the rules and work within the system as it is.  But, there will always be a contingent looking to beat that system by going outside its boundaries and, with each innovation in the ad marketplace, we need to keep an eye out for a new spin on the same old tricks.

Edit 4/27/12: Added clarification to first paragraph specifying that we are referring to cases where the affiliate usurps the display URL of the company.

  • Mark Ballard
    Mark Ballard is Director of Research at RKG.
  • Mia Brennan
    Mia Brennan is Director of PPC Testing & Analytics at RKG....
  • Comments
    6 Responses to “More Affiliate Dirty Tricks: Device Targeting Brand PPC”
    1. Terry Whalen says:

      Thanks for the article. We have an ecommerce PPC client and we found that an affiliate has been bidding successfully on a core brand keyword for possibly years. Interestingly, we increased our bid to $100 and then $200 for the brand term, thinking that the actual advertiser account would be rewarded with a higher ad rank than the affiliate account, at such a high bid. But we have found that no matter how high we bid, the affiliate wins the auction. So, time to track down the affiliate.

    2. Mark Ballard Mark Ballard says:

      Yeah Terry, in most cases the only real solution is to track down and boot the affiliate from the program. It’s pretty amazing that Google can’t or won’t take steps to prevent this themselves and, while we likely won’t get caught paying $100-200 per click, having to increase bids to those levels carries risks and can drive up actual CPCs above what we really need to pay. Unfortunately, tracking down the specific affiliate isn’t always a simple effort given the layers of obfuscation some employ.

    3. Terry, one challenge with outranking affiliates is that they likely have a HIGHER CTR than the brand does because their copy can read “Save X% at [Brand}” which is a very compelling message for users, and often NOT the message the brand wants to deliver on navigational traffic.

    4. Jeremy Brown says:

      Previously, Exact Match Impression Share for well-structured brand campaigns could be a good proxy way of spotting affiliates on brand terms.

      However, Google has changed the way they calculate Exact Match Impression Share and it’s now done by grouping up any instances of the display URL.

      I’ve been communicating with Google why this is a problem (see below), but it would be great if others would pass along their thoughts to their contacts at Google as well.

      1. Adwords is now reporting inaccurate data.

      The interface shows the number of impressions, but the Exact Match Impression Share does not correspond to those impressions. Rather it corresponds to some unreported number of impressions that are totaled using data from other Adwords accounts that cannot be accessed.

      This inaccurate data makes it easier for affiliates to mask the poaching of brand terms. I highly recommend Google switch back to reporting data for one account and not some unknown amalgamation of accounts. Also, that change should be made retroactively. If that’s not possible, then I recommend reporting those other impressions and their sources; provide a report that gives a breakdown of the different accounts contributing to your Exact Match IS.

      2. There was zero notice of this change.

      Yes, there was an Inside Adwords post broadly discussing changes to IS metrics, but there were no details mentioning that metrics would now be calculated using the display URL and possibly from a number of Adwords accounts. Google needs to provide greater transparency around sizable changes like this.

      Further, the documentation around Exact Match Impression Share is either incomplete or misleading, as there is no mention of display URL. This is a problem in general.
      http://support.google.com/adwords/bin/answer.py?hl=en&answer=52760

    5. Mark Ballard Mark Ballard says:

      Good points, Jeremy. With many advertisers utilizing multiple accounts for the same program, Google needs to make the connection through some method, but this is a sloppy means to do so. The core problem remains Google not having a robust validation mechanism in place to determine who is really who and who has authorization to use a given display URL.

    6. Terry Whalen says:

      Jeremy, that was news to me. Thanks for the heads-up. I had been wondering about this very question lately.

      George, good point. In our case, the affiliate ad doesn’t seem to have any obvious attributes that would result in higher CTR, but I can see how your scenario might happen quite a lot.