Abonnez-vous au flux RSS (français) - -
Subscribe to the RSS feed (english)
In the war that opposes FaceBook to MySpace, there is another war, the audience metrics war.
MySpace is open, accessible to all.
FaceBook is closed, you must be logged to access it.
MySpace's audience is measured in unique visitors.
FaceBook's audience is measured in members.
There are obviously fewer members than visitors, because a member who is using several computers corresponds to several visitors, even "unique" ones. You could object that there can be several members on a given computer (in a household or in a cybercafé), but on average, uniques visitors are more important.
So, what metrics is closer to reality?
The concept of member is a better estimation of what I've called in a previous article, the "idiot" visitor.
At one condition: he/she must have good reason to identify. This si the case for FaceBook, but it quite rare that 100% of visitors log-in (and some Google services for instance). Most of the time, only a part of the visitors will identify. But as soon as a visitor is identified, why not use this information? Let me make myself clear... Alenty's audience measurement system can recognize when visitors are identified on a website. Those who are not, are measured, as all site-centric tools, by a cooke. So, we can do the complete usage analysis, using either the member identification (when available), or the cookie. This way we get a better view of the real site audience.
Maybe fewer visitor numbers (because they have been deduplicated), but a better estimate of their visit frequency, the time they spend on the site, and the possibility to link these members to the internal website clients database! All this, of course, automatically, and transparently.... What do you think of such a mix model?
At one condition: he/she must have good reason to identify. This si the case for FaceBook, but it quite rare that 100% of visitors log-in (and some Google services for instance). Most of the time, only a part of the visitors will identify. But as soon as a visitor is identified, why not use this information? Let me make myself clear... Alenty's audience measurement system can recognize when visitors are identified on a website. Those who are not, are measured, as all site-centric tools, by a cooke. So, we can do the complete usage analysis, using either the member identification (when available), or the cookie. This way we get a better view of the real site audience.
Maybe fewer visitor numbers (because they have been deduplicated), but a better estimate of their visit frequency, the time they spend on the site, and the possibility to link these members to the internal website clients database! All this, of course, automatically, and transparently.... What do you think of such a mix model?