In preparation for Web 2.0 next week, I spoke today to Gian Fulgoni, founder of Comscore. He’s giving a great talk on the things a major research house knows about the web. One of the slides he showed me I just had to steal to give you a sneak preview (click on image for a larger view).
What this tells us is pretty interesting: the heaviest users of search, who are a minority of total search users, account for the vast majority of search queries. This seems a case where the tail is not as powerful as the head….And clearly, the more folks use the web, the more they use search. What happens when the majority of us are heavy users of search? Time to buy more servers….
As you note the table suggests the more active a user becomes the more he or she searches. This contradicts the actions of short sellers of internet search companies as you noted in an earlier post today.
Short sellers don’t dig deep enough.
Isn’t this just a case of the 80:20 rule:
80% of your time & energy / sales / whatever will come from 20% of your tasks / customers / whatever.
Presently the top 20% of users only account for 69% of all searches. Following the 80:20 rule, heavy users are not doing enough searches!
So if anyone did want to know why personilisation was being persued faster that buck rodgers, you now have the answer. Raise switching costs, increase your ARPU (sponsored links) and your targets should be achieved by users naturally going up a usage curve.
These are the very same metrics why “triple play” is so important as a integrated offering (voice, data, TV) All delivered of course over IP from “a platform” that sits on “the network”
It would be interesting to see what type of user is responsible for the most clicks on advertisements. Do the people who are doing the most searches generate the most revenue? Or are they so familiar with the placement of advertising that they ignore it all together? Are light searchers, more likely to confuse advertisements with search results and as a results generate more click-throughs?
Teresa
Hey John, did Gian mention what the time-period is for the searches-per-searcher statistics? Weekly, monthly?
It’s monthly