Web Traffic Numbers Down? Don't Count On It!

  • Written By: D. Geller
  • Published: October 22 1999

Event Summary

The respected Internet and Digital Media measurement firm Media Metrix announced its monthly survey of traffic on the "Top Fifty At Home and At Work" web site traffic ratings. The listings show traffic reports for individual web sites and for "properties." A property represents all of the sites belonging to a particular company like Yahoo, AOL or Microsoft. Figure 1 shows the summary measurements from May through September.

Figure 2 looks at the number of visitors in 1998 and 1999 for the same months. (Missing data points were not available from Media Metrix).

Figure 3 shows unique visitors segmented by use from home and use from work; for these Media Metrix reports only the top 25 sites.


What is most important to note is that these numbers are limited snapshots of a very complex stream of data. They are certainly valuable for understanding long-term trends, but small fluctuations should not be overemphasized. Where Figure 1 seems to show the precipitous drop being reported anxiously in various headlines, Figure 3, which shows a slightly smaller data set, reduces the anxiety level considerably. In fact, recasting Figure 1 with the bottom axis at the zero point also gives a less terrifying view (see Figure 4).

Does this mean that there was not a drop in usage for September? The drop, as Media Metrix points out in its own press releases, had a lot to do with the hurricane season. Visitors to top weather sites such as Weather.com and NOAA.gov rose 21 percent over all (see Figure 5). There were similar percentage increases in visits to top sports and news sites

User Recommendations

Do not change your E-commerce plans based on these news reports. There is no reason from these data to have any concern about the Web reaching saturation in any way. The perceived decline shown in Figure 1 is most likely a combination of external events (hurricanes, playoff races, earthquakes in Turkey) and generally good summer weather. If there were a hypothesis to be tested over the next few months it would be a shift from the big portal sites to more specialized content. We expect (probability 70 percent) to see small swings in the balance between portals and more focused sites (including vertical portals) over the next 18 months.These are unlikely to pose much of a threat to the major portals and Web publishers.

comments powered by Disqus