Yahoo! Slams comScore

From Yahoo's (YHOO) earnings call transcript; Yahoo President Susan Decker is speaking about Yahoo's metrics and comScore (SCOR):

...when looking at third-party services such as comScore to assess unique users or time spent the aggregated figures may not tell the story of what’s happening and the key, value-creating starting points for consumers and advertisers. Our internal logs show that the metrics we’ve discussed with you in the past, such as uniques and page views, continue to grow in the double-digits in Q4 with unique users now topping 500 million and page views about 4 billion per day.

With consumers accessing the web in so many ways, we’ve looked for a more unifying global metric that’s more flexible across Yahoo!’s and our partners’ properties and useful across multiple devices and geographies. We expect to use visits to Yahoo!’s global starting points and anchor sites to be the most relevant metric going forward.

Visits are defined by comScore as the number of times a unique person accesses content within an entity and they capture the fact that consumers may visit a property multiple times a day from different locations or devices; not just whether they touched a site once per month.

Larry Dignan writes:

On the conference call there were a few key takeaways that are worth noting. To wit: ...Decker kicks Comscore. Decker noted that Yahoo’s page growth was growing at a double digit clip in contrast to what Comscore is reporting. Welcome to the club Sue. Decker noted that the company is using visits instead of page views and unique users as a more appropriate metric for Yahoo’s key properties.

Here's the full YHOO earnings call transcript.

Comments

[Balanced_View]

Full disclosure: I worked for comScore years ago, and as a result I do own some of the stock. But I have not been a comScore employee for years, nor does my livelihood by any means depend on the performance of the company.

Seriously, do Decker's comments "slam" comScore as your headline suggests? Dictionary.com defines "slam" in this context as "to criticize harshly; attack verbally." I don't see that here, I see a contrasting view of what different numbers say, and a disagreement over which to use for different applications. (By the way, no question the ZDnet blog post is a slam.)

But the public audience metrics cited publicly, and which sometimes draw disagreement, are only the tip of the iceberg of the information companies like Yahoo! and 1,000+ others receive from comScore and Nielsen about consumer behavior, intentions etc. The vast majority of the information these companies pay for is never published and/or doesn’t attract these debates.

Why have so many companies continued to buy this stuff for almost a decade? Why did so many buy it well before it was so widely published? Because for the vast majority of businesses, it works. The extent to which it is valuable far outweighs the number of situations in which it is not.

Could it be because there is no other game in town? That can't be, because there are plenty of alternatives, and history documents that many more companies gave this business a shot but failed to deliver enough value, accuracy or whatever to sustain themselves.

My experience tells me that those who go to critical extremes on this issue are a vocal minority with any number of motivations. Of course the numbers are not all perfect, show me something that is. A balanced view of this situation is appropriate. But then, dramatic words in bold print are usually flashier than balance.

[DavidJackson]

Balanced View, I think you're right: the headline does suggest harsher criticism than Sue Decker's tone suggested, as opposed to Larry Dignan's comments. Thank you for calling this out.

At the same time, if major players like Yahoo and ZDNet/CNET say that comScore's numbers no longer accurately reflect their businesses, that's got to be worth taking note of.

Question for you: how much of a threat to comScore are companies like Compete and Quantcast that make a large amount of data publicly available for free?

[Balanced_View]

I'll preface this by saying I've been out of the business for a long time, so I'm not as close to the details of what each of these companies is doing now.

Compete has a bunch of clients and probably some very smart people, and has been around just about as long as comScore but has never achieved the same traction, broad credibility or scale. The company has always seemed to have its eggs in a lot of baskets. It started out going head-on against comScore. Then it focused (seemingly well) on just a couple of specific vertical markets. Then free data to the world. Then target-marketing services. It is unclear how big, profitable and complementary each of these products are.

I'm not familiar enough with Quantcast to comment in great detail. As best as I can tell part of its strategy is to exploit the very grumbling that prompted this thread, by combining site server data with some kind of panel. Companies that don't like panel data usually point to disagreement with their server data as the basis for inaccuracy. But the problem is that server data have been riddled with their own problems since the beginning of time (well known). Perhaps Quantcast has some interesting new way to calibrate these sources.

There are other significant comScore capabilities that differentiate the company. One example is the ability to measure actual transactions including purchases, applications etc. Other companies including Nielsen have claimed over the years to have their own abilities in this arena, though I don't know how much commercial progress any of them have made. I do know that one should not dismiss the potential of a determined and well-funded competitor, of which there are a couple. At the end of the day comScore's track record, which is now public record, speaks for itself.

There is a challenging sweet spot of many variables including market relevance, accuracy, industry credibility, price/value, audience and outlet coverage, scalability and (of course) profitability/investor return that a company needs to find to be successful in this business. To date few have pulled it off.

[DavidJackson]

Very interesting and helpful -- thank you!

One follow-up question: do you see any possibility that Google Analytics, which is powerful and free, could offer an option to publishers to make a limited amount of data public, thereby offering a more accurate and consistent yardstick for mid-sized and smaller websites?

[Balanced_View]

In theory anything is possible, but I think this unlikely. Anyone can quote their raw server-based uniques, page views etc. to the world, and that's been done for years regardless of the software used to get them (Omniture, Coremetrics, Google Analytics etc.). It doesn't usually come across as clean, objective data, even if it's right off the box.

Ad buyers typically also want demographics as well, which these sources may not have accurately or at all.

To make the info more convincing it needs to be processed, reviewed, cleaned, and maybe projected, calibrated, etc. etc. by a reasonably objective third party. And include demos. Looks like Quantcast is trying this, and others (such as I/PRO) have tried it before. All of this requires a lot of work, and in the past there haven't been enough small publishers and advertisers willing to pay enough to make this model profitable.

If an analytics software co. publishes the information, then as soon as the numbers go south for a site, that site will suddenly be much less interested in participating in (or paying for) a program that does not let them control the release of their numbers. Plus someone has to answer the calls and emails from publishers complaining that their numbers weren’t right because tags weren’t placed right, their affluent visitors don’t answer demographic surveys, they redesigned their site etc. etc.

That puts a real damper on trying to be an objective third party. As the saying goes, it’s all fun and games until someone loses an eye.

There is the whole other issue of publishers being uncomfortable with Google being the arbiter of this information. Not enough time for that topic.

For these and other reasons I would guess this not to be an appealing business to Google.

[AndrewLipsman,comScore]

While we stand by comScore’s data as representing trends in visitation to Yahoo! Sites by people using computers around the world, it is certainly possible that Yahoo! Is experiencing growth in pages downloaded by Internet users who are also visiting Yahoo! sites from mobile devices or locations such as Internet cafes, both of which are not currently included in the comScore sampling definition. In fact, given that Yahoo! now has branded Internet cafes in several developing markets, it’s quite possible that Yahoo! properties attract particularly strong visitation among these users.

That said, we would caution against drawing conclusions regarding increases in unique visitors if those conclusions are being based on site server log data that use cookies as a means of counting unique visitors. In that case, it is entirely possible that a visitor whose visitation is increasing through multiple access means (e.g. using a mobile device in addition to a computer) will be counted as two visitors when in reality it is only one.

With regard to page view metrics, the page view has been rendered less relevant by interactive technologies such as those used in Yahoo's beta mail product. Such technologies issue many more web calls to render the end-user experience. In these environments it becomes very difficult to "count pages" in a manner that is consistent with legacy static-page environments. This difficulty exists in internal site server logging as well, making comparisons between internal and third party measures even more complicated.

[DavidJackson]

Andrew, it's good to have your input to the discussion. Thank you for participating in Seeking Alpha.