Four take-aways from the ComScore May traffic numbers (ASKJ, CNET, GOOG, IACI, MSFT, NYT, TWX, YHOO)

|
Includes: AABA, ASKJ, CNET, GOOG, IAC, MSFT, NYT, TWX
by: David_ Edwards

From American Technology Research analyst David Edwards note to clients (relevant to investors holding positions in ASKJ, CNET, GOOG, IACI, MSFT, NYT, TWX, YHOO):

Looking at Internet Traffic Patterns

comScore released its May 2005 Media Metrix results yesterday.  ComScore provides three metrics that we find particularly useful for gauging both the quantity and quality of traffic to various Web Properties: 1) Total Unique Visitors, i.e. the quantity of raw, unduplicated traffic to a property; 2) Percent Reach, i.e. the percentage of the total U.S. Internet user base that visits a particular property; and 3) Total Pages Viewed, which is essentially a measure of how much "inventory" is viewed on a given property. Taken together, these three metrics provide a solid means of determining the overall quality and viability of a Web Property.

Our four take-aways from the data:

1. Yahoo!, Time Warner and MSN remain the top three sites across all three metrics.

2. Although Google is not far behind in number of users, its number of pages viewed is significantly lower (note, we explain below why we do not necessarily see this as a negative).

3. Both AskJeeves and CNET Networks are showing strong growth across all metrics...of course, some of this growth is inorganic.

4. We find it interesting that there are three "old media" companies (i.e. they've been in business for more than 10 years) which have transformed their businesses enough to hold a spot on the Top 10 list -- Time Warner, Viacom, and The New York Times.

Note that we believe that any conclusions drawn from this data should be handled cautiously. The two biggest caveats are 1) the data only covers U.S. Internet traffic, and 2) the data does not include conversion rates -- e.g. the conversion of traffic into sales. So we're hesitant to draw P&L conclusions from these results. And we're much more likely to lean on Y/Y trends rather than M/M trends.

Full disclosure: at the time of writing I'm short CNET.