By Danny Sullivan, Editor
The Search Engine Update, May 13, 2002
This sidebar article compares key metrics about search engine popularity and usage, as provided by the web’s major rating services, Jupiter Media Metrix and Nielsen//NetRatings. To best understand it, you should have previously read the Google Tops In “Search Hours” Ratings article.
Why would these services have different results? There are a variety of reasons, with one of the key reasons being how “search specific traffic” is defined. Each service has to decide which domains will be counted toward a particular company’s totals.
When you have a company running multiple web sites, or distributing results through affiliates, it is more likely that the two measuring services will differ in how they calculate the company’s traffic.
Current Audience Reach
The first chart below compares audience reach estimate from both services, for March 2002. The lines shown would perfectly overlap, if the two services were in exact agreement over the reach of the search engines. Alternatively, if the lines were separated but identical in basic trend shown, then the services would at least agree on the relative popularity of each search engine. Anywhere there are big divergences, this is where the two measuring services are in significant disagreement about a particular search engine’s popularity.
US Web Audience Reach
Home & Work Users
March 2002
KEY: YH=Yahoo, MSN=MSN, GG=Google, AOL=AOL, IS=InfoSpace,
AV=AltaVista, OVR=Overture, LS=LookSmart, NS=Netscape
NR=NetRatings, JMM=Jupiter Media Metrix
As you can see, the big disagreement is over the popularity of Yahoo and MSN. NetRatings puts Yahoo first, while the Jupiter line starts low, then shoots up because it sees MSN as more popular.
Both services agree that Google, AOL and Ask Jeeves are the next most popular services, then it’s back to disagreement over InfoSpace. NetRatings ranks the company sixth, but Jupiter would put LookSmart into that spot.
Average Minutes Searching
The next chart compares the average time spent per visitor at each site, for March 2002:
Average Minutes Spent Searching Per Visitor
Home & Work Users
March 2002
KEY: As above, except DP=Dogpile
Both services agree that Google gets the most time spent searching per visitor, per month. NetRatings didn’t report to me a time figure for AltaVista in March 2002, which is why there is a gap in its line. Both then agree that Ask Jeeves is the next behind Google. After that, there are plenty of differences of opinion.
Average Minutes Searching
This last chart compares search hours. Unfortunately, Jupiter Media Metrix only provided the number of unique visitors to each site for March 2002, not also the average amount of time spent at each site by those visitors, in that month.
To make up for this, I used the average time figure for each site from Jupiter for February 2002 multiple by March 2002 unique visitor counts. This alone could cause differences between the NetRatings line, which is entirely based on March 2002 figures. However, it may still provide a useful rough comparison.
Total Search Hours
(Millions of Hours)
March 2002
KEY: As above, except DP=Dogpile & DH=Direct Hit
Again, there is agreement on Google but differences beyond that. Indeed, the differences between Jupiter and NetRatings over visitors to Ask Jeeves and the time spent by those visitors at the site get dramatically magnified when both are multiplied to get search hours. Suddenly, Jupiter sees Ask Jeeves as more used than MSN Search.