Google Q1 Results Challenge comScore

Publishers and site owners have long complained that third-party traffic measurement services such as Nielsen and comScore under-count their traffic and unique users. For example, Yelp sent me an email in late Feb saying that internally it shows more than 8 million monthly uniques, which was more than some of the third party services reported. See this early March comparison of Yelp, Citysearch and Local.com (per Compete):

compete march

I just a moment ago looked on Compete and, based on the earlier email, these numbers would now appear to correspond to the internal counting and trajectory of the Yelp numbers (not sure re Citysearch and Local.com).

compete data

According to the Wall Street Journal, comScore is taking some heat (and from investors) over apparent discrepancies between its metrics regarding declining paid clicks on Google and Google’s strong Q1 results. comScore itself tried to clarify there wasn’t necessarily a direct link between its data and Google’s performance. But Wall Street analysts, with some justification, continued to obsess on the paid click data.

(Google said in the earnings release, however, that paid clicks grew “approximately 4% over the fourth quarter of 2007.”)

I never directly reviewed the recent, controversial comScore reports themselves but have read the data in third party discussions. For its part comScore appears to have been incorrect on the raw numbers and was perhaps mistaken to not put enough caveats around its data. However, financial analysts were insufficiently sophisticated about the metrics and user behavior to be more “restrained” and “sober” about it in their corresponding “notes.”

There’s also a herd mentality that exists in such situations; and there’s some segment of the population that would have liked to see Google stumble and be proven an “ordinary company” not immune from the pressures facing others. That created a certain receptiveness to the declining clicks information.

However, I’ve always felt that the focus on paid clicks was misplaced as an indicator of whether Google was suffering from the larger problems of the economy. The better metrics would have been query volume and market share (comScore discusses this to some degree in the “clarify” link above). Paid clicks is a function of ad coverage and search volume. If search volumes had declined (which they haven’t) or Google had been losing market share (which it hasn’t) then these concerns would have been more justified.

Because there’s no “cost” to consumers to click on ads, clicks are not necessarily going to suffer in a recession. What will likely suffer are certain kinds of commercial queries. For example, if I put off buying the new car then I might not be doing as many searches for cars. That in turn will translate into fewer paid clicks. However, in the aggregate, query volumes were not going down (as comScore said):

The most puzzling data element is that Google’s U.S. paid clicks dropped sequentially by 7%, while, at the same time, its total number of search queries grew by 9%. At the same time, Google’s market share of all search queries grew slightly from December, and its annual query growth remains very strong. All indicators point to the company continuing to do very well as far as consumer usage and competitive position.

Some questions are approrpiate to raise about whether comScore’s methodology needs some fixing. Yet there should be no schadenfreude. Third party measurement is critical to industry credibility. Simultaneously, everybody has to take all the numbers, whatever their source, with some perspective and see them not as literal truth but rather as directional indicators.

You can follow any responses to this entry through the comments feed.

3 Responses to “Google Q1 Results Challenge comScore”

  1. Michael Stoppelman says at

    Greg,

    I think the comScore numbers might not accurately reflect Google’s number because international traffic is so important to Google. I would suspect most of the 4% of growth was mostly from outside the US. Google states publicly that over 50% of their revenue is non-domestic. The question to ask comScore is how well they measure international traffic. I can’t imagine they can get ISP logs from China. Maybe I’m wrong :).

    Disclaimer: I worked for Google as a software engineer on AdWords/AdSense (2003-2007). I currently work at Yelp as a software engineer.

  2. Scott says at

    Google’s paid click growth, which I believe was up 20% y/y, includes both US and International. ComScore is only US data so comparing the two numbers needs to be done with this perspective in mind. Also, Google’s US y/y revenue growth went from 40% in 4Q to 29% in 1Q08. That is the largest rate of decline in its history. It’s still large growth on a large number, but would give me some pause as an investor. Keep in mind that Google included some revenue from DCLK this quarter and has done a number of acquisitions, the largest of which is YouTube, which added a bump to growth and suggests that Google US Search growth was probably closer to the 25% y/y in 1Q. Thats still a far cry from comScore’s number for paid clicks, but with the improvement in quality, it’s not crazy to think a lot of the growth could have come from higher pricing rather than more paid clicks.

  3. Greg Sterling says at

    Agree with the international statement and, Scott, with your suggestion that directionally it was correct. My understanding, however, is that click prices are stable or flat.

Leave a Reply