Google v Facebook: Lies, Damn Lies and Statistics

I’ve come to the conclusion in the last few weeks that some bloggers and journalists just shouldn’t be allowed near numbers.  They clearly don’t have any sort of grasp of statistics or how to do a comparison.  Their memory or knowledge of the history of their subject is woefully inadequate.  However by thinking about how they misunderstand statistics we can understand our own websites better.

What they do have is a headline in mind and they try to force the numbers to fit.  Here are a three of the variations I’ve seen:

Google+ Fails to beat Facebook – the argument here goes Facebook has 700 billion trillion gazillion users and Google+ has only 20 million.  At no point does the article point out that Google+ launched only weeks ago and is still in beta.

Google Fails because Facebook has more time on site per visit or pageviews – here the mistake is comparing Apples with Oranges.  They’re not comparing Google+ with Facebook – they’re comparing Google (primarily search) with Facebook.  If the time on site for Google was the same as Facebook I’d be concerned.  The point of a search engine is to get you to the good information quickly.  The point of a social network is to interact with your community.  These are two different tasks and the meaningful metrics are radically different.  Once we know how Google+ stacks up against Facebook this would be a valid comparison.

Google+ is just for men - All new social networks tend to be picked up by certain groups ahead of other groups.  Network analysis of social networks has shown that people with an interest tend to invite people with the same interest.  They have also shown that many interests have either lots of men or lots of women involved.  One early mailing list system found that for the first few months most of the lists set up were created by people who were interested in lizards.  Every day new lists for lizard lovers of different types would be added.  Then some amphibian and bird themed lists crept in.  Then they reached pet lovers and at that point lists for all kinds of interests appeared as people with pets and other interests signed up.  Social networks take time to grow and diversify.

While reading about Google+, Google, Facebook, Yahoo, Bing, Apple, Microsoft and all the other companies and services beloved of pundits ask the question: Do the Numbers Make Sense?  What would good numbers for Facebook or Google be?

Then think about your site.  What would good numbers for it be?  Are there factors that skew the numbers just like there are factors affecting the big players?  Can you alter some of those factors or do you just have to understand them?  Is a number bad or good in your context?  As otherwise all the numbers are just Lies, Damn Lies and Statistics.

Authored by Ian Cockayne