Silicon Investor (SI) -- The First Internet Community

STOCKTALK

We've detected that you're using an ad content blocking browser plug-in or feature. Ads provide a critical source of revenue to the continued operation of Silicon Investor. We ask that you disable ad blocking while on Silicon Investor in the best interests of our community. If you are not using an ad blocker but are still receiving this message, make sure your browser's tracking protection is set to the 'standard' level.

Technology Stocks : Netscape -- Giant Killer or Flash in the Pan? -- Ignore unavailable to you. Want to Upgrade?

To: Kal who wrote (4169)	8/26/1998 9:19:00 AM
From: Ted Shelton	Read Replies (1) \| Respond to of 4903

"Hot 100" has a methodology which is substantially inferior to audience measurement companies Relevant Knowledge and Media Metrix. Hot 100 measures traffic at certain points on the network, with no attempt at making sure that these points are representative of the web surfer audience universe. Here is what Hot100 says about their own methodology: We primarily use the logs from proxy servers. These logs are FTP'd to Web21 on a daily basis and represent the surfing patterns of over 100,000 surfers world-wide. We protect the privacy of the people we sample, so the exact demographics are not known. Approximately 60% of the people we sample are from North America and 40% are from international countries. Universities, businesses, and home users are all represented, but AOL, Prodigy, and Compuserve users are not represented. more details are available at: hot100.com One advantage of the hot100 system is that it is gathering a sample from both the US population and a worldwide population. From this perspective it may be true that Netscape has the most visited web site. However the hot100 methodology cannot be viewed as reliable. In contrast to this ad hoc data collection method, Relevant Knowledge and Media Metrix each employ a team of statisticians that have been, over the past few years, building a model of the web universe as a whole, and creating a "panel" which is a statistically representative sample of that web user universe. This is an established and accepted methodology -- examples in other mediums include Arbitron in radio and Nielsen in television. Measurement of the web audience size and composition, determination of a representative panel, and recruitment of that panel all follow careful scientific methods. For more details on these companies see: rkinc.com and mediametrix.com (sorry I couldn't find the specific place on their site where they describe their methodology) Another contender in this space that someone may wonder about, so I will address this for completeness, is NetRatings. This company is a newcomer in the audience measurement space and has received a fair amount of skepticism about the validity of its data because of its panel recruitment strategy. Initially when NetRatings got started they used banner advertising to attract online surfers to sign up as panel members. This would appear to create a panel skew toward certain sites (ones the banners ran on) and toward more active web surfers. In general the industry considers Relevant Knowledge and Media Metrix to be the preferred data providers. FYI - Netscape does not rank #1 on either poll. Here is Media Metrix top 10 from home: aol.com 46.0 yahoo.com 40.5 microsoft.com 27.3 geocities.com 25.9 netscape.com 25.4 excite.com 23.2 infoseek.com 15.6 angelfire.com 15.1 lycos.com 14.5 The numbers after the domain name represent the %age of the total projected US web audience that each of these domains reaches. Note that this does not include domain consolidations but just individual domains -- so Angelfire is now owned by Lycos and, in the consolidated report, would be included. The numbers are not additive though, as they represent a unique audience, and there will tend to be overlap between sites (e.g. some users have visited both). When looked at from the workplace, though, Netscape does much better -- yahoo.com 49.0 netscape.com 40.0 aol.com 34.1 microsoft.com 33.0 excite.com 30.3 infoseek.com 22.8 geocities.com 22.1 AltaVista Search Services* 19.9 lycos.com 19.7 msn.com 14.3 In the case of the "at home" measurement, the statistical margin for error is +/- 1%. The "at work" measurement is much less precise, with a margin of error of +/- 3%. This larger margin of error represents two issues, first a much smaller panel size against the total universe of at work users and second a much shorter period of operation during which the methodologies for measuring this audience may be refined. All of this data is freely available on the Media Metrix web site and similar data is available on Relevant Knowledge. yours, Ted Shelton