Top online brands and the calculation of charts

My friends at Futurelab have posted their own chart of the Top 100 Online Brands, as an alternative to that provided by Interbrand (here’s Interbrand’s 02005 list, with the 02006 one due soon [Update, 28 July: the link to 02005 list now redirects to the 02006 list]).
What does a chart of brands mean, and to whom? I can see that the branding industry wants to identify who’s doing best, and then determine how they’ve achieved that. But what can the rest of us learn from the results that Kodak beats Coca-Cola on Futurelab’s metrics, but Coca-Cola was top of Interbrand’s list? For most of us, these two brands do not overlap or compete in any part of our lives.


This is purely subjective, but I did notice that seven out of Futurelab’s top ten were brands that I’ve actually come into contact with a reasonable amount (more than ten times, say) in the last year. On the other hand, I couldn’t find more than five or six in the remaining 90 that I think I might have used more than ten times. I suppose this means that, offline, I live a bit of a No Logo lifestyle.
Futurelab admit that the methodology behind their Top 100 is pretty rough. One aspect of it intrigued me for its apparent un-European-ness (Futurelab are based in Belgium): they counted the number of times the wording “I hate (brand)” and “(brand) sucks” appeared in Google. Surely, on this side of the Atlantic, we have many alternative (and, dare I say, richer) terms to express derision than saying that something “sucks”?
We need to pick carefully what we choose to chart, for whom, and find the most sophisticated measures we can to chart it. What I’d like to see would be first some way of clustering brands — perhaps like Flickr clusters its tags, so that you can compare BMW with Mercedes with Ford, and Hewlett-Packard with Dell. Then within each cluster, perhaps some hybrid measure of use and ratings, again like Flickr’s interestingness.

One thought on “Top online brands and the calculation of charts

  1. Hi David
    Thanks for your thoughts and suggestion on the clustering. We’re waiting for the Interbrand 2006 to come out before doing a recalculation, yet I was thinking to group by category or (as you say) cluster. Your comment clinches it.
    In this context, we may actually extend the list beyond those Interbrand lists (as there are brands which are not even featured in the Interbrand list who are extremely relevant online).
    As for the use of language, I fully agree with your point and have debated whether to use s#8ks as a parameter or not (every time I read it, it bugs me).
    However, in contrast to many other variations of the theme we tried (European and American), including or excluding this one does significantly influence rankings, mainly when it comes to US brands.
    Perhaps in stead of begging pardon for our “french” we should say “pardon our american” :-)
    Alain

Leave a Reply to Alain Thys Cancel reply

Your email address will not be published. Required fields are marked *