Last week, eBay pulled all of its Pay-Per-Click ads from Google in the United States. This week, Hitwise reported that eBay’s traffic actually rose 0.9 percent in the week after its campaigns were shut down, and that its traffic from Google declined by just 0.9 percent.
What’s wrong with this picture? Let’s run some numbers and do some digging.
EBay is the largest single advertiser on Google and spends about $25 million each quarter on PPC ads. Its in-house search team buys millions of keywords and makes extensive use of Dynamic Keyword Insertion to capture “long tail” queries. Before eBay withdrew its ads last week, you could search for almost anything on Google and eBay would provide a sponsored listing, a situation that often produced amusingly absurd results (“Looking for open wounds? Find them on eBay”).
The long tail keywords that eBay buys aren’t worth much: probably just a few pennies a click, but let’s assign an arbitrary value of $0.15. $25 million spent on $0.15 clicks adds up to 166 million clicks per quarter, or 12,820,512 clicks per week.
If Hitwise’s traffic numbers are correct, this means that eBay, after turning off all of its Google PPC campaigns last week, was somehow able to generate almost 13 million clicks in addition to its ordinary traffic from other sources. Where did those clicks come from? Nobody seems to know. Did those clicks result in a comparable conversion ratio for eBay’s many merchants? Again, nobody outside of eBay knows.
Frankly, I’m highly doubtful that Hitwise’s numbers are on the mark. It’s inconceivable to me that eBay, after turning off the biggest PPC campaign on the planet, wouldn’t see a massive drop in traffic from Google. Even without access to eBay’s internal traffic and conversion data (Did-it doesn’t manage eBay’s search campaigns), it’s obvious that doing this would create the traffic equivalent of a Mt. St. Helens-style event. In my view, the fact that Hitwise didn’t pick it up says more about Hitwise’s methodology than it does about Google or eBay. If your seismograph doesn’t pick up Mt. St. Helens, don’t blame the volcano: blame the machine that missed it.
Hitwise’s methodology (which uses ISP sniffing technology to establish its traffic numbers) has been roundly criticized in the past. Here’s Google’s Matt Cutts commenting on his Blog about the reporting flaws which can crop up in Hitwise’s ISP-sniffing approach to Web metrics:
ISP relationships can be a huge source of metrics bias. For example, some ISPs partner with Yahoo, and users on those ISPs are probably more likely to visit Yahoo. Other ISPs partner with Google. And savvy users that use smaller providers such as Covad or Speakeasy are likely not counted at all. Because you don’t know which ISPs are selling user data to companies such as Compete or Hitwise, you don’t know what biases are baked into those companies’ metrics-and the metrics companies won’t tell you.
Just for the sake of argument, let’s assume that Hitwise’s numbers are right, and that eBay plucked those 13 million missing clicks out of thin air. The ramifications of this are stupendously troubling not just for Google, but for eBay. Over the past five years, I’d estimate that eBay has spent nearly a half-billion dollars on Google PPC ads. If Hitwise’s numbers are accurate, eBay’s current management, including its in-house search team, should be fired tomorrow because it clearly didn’t perform any tests on the relationship between PPC spend, traffic and merchant conversions anytime in the past five years. Furthermore, eBay’s shareholders should immediately institute a lawsuit alleging a gross breach of management’s fiduciary responsibility. Before tossing the whole team on the sidewalk, however, I’d ask Hitwise to do a lot more explaining about its findings than its spokesman have done so far.
We’re in a very peculiar, very weird situation in this industry. The whole promise of Internet marketing is predicated on transparency and accountability. Journalists, investors and marketers depend on numbers produced by metrics companies, but these companies provide less than an optimum level of transparency and accountability as to the accuracy of their data. Tens of billions of dollars are at stake here, but there’s absolutely no guarantee that the numbers upon which such decisions are based have a verifiable correlation to reality. Instead, the whole industry is being guided by “solid” metrics which, like the Maltese Falcon, are little more than “the stuff that dreams are made of.”