This article at WebProNews falls into the “Lies, Damned Lies and Statistics” trap by suggesting that MSN delivers far fewer buyers than does Google. That being the case, since Google delivers 65% of the search engine referred traffic to the vast majority of web sites and MSN only 15%, their statistically higher percentage of buyers is far less significant to the numbers of visitors delivered.
Let’s turn 100% into 100 people, OK? Given 48% of 15 people (being generous with estimates of MSN sending 15% of traffic) that is 7 people ready to buy. Google on the other hand delivers 65 people (percent) and 42% of them ready to buy something – THAT IS 27 people who came from Google “Ready to buy” things versus 7 people from MSN “Ready to buy” things. MSN may have higher percentages, but FAR lower numbers to draw that percentage FROM.
This shows clearly that Google sends 4 times the number of buyers as MSN does. That is if there really are many ecommerce sites that receive 15% of their referred search engine traffic from MSN. Most studies show that number as being much lower in actuality.
This percentage game is an absurd numbers game ComScore MediaMetrics plays to make MSN & Yahoo look good every quarter as we’re discussing financial predictions and it makes me crazy. Every news organization dutifully reports the percentages and perpetuates the lie that Yahoo and MSN are doing better than Google. It is a complete fabrication.
Let’s rate AskJeeves Search against Google. Since AskJeeves sends 1 visitor for every 5000 that Google sends, would it matter if 100 percent of AskJeeves searchers ALWAYS purchased something? No.
If only 1 percent of Google’s 5000 visitors purchase something, then they still beat AskJeeves by 50 times, even though their “intention to buy” statistic is (in this absurd example I’ve given here) only 1 in a hundred. It matters far more that Google sends 500 times the traffic as AskJeeves than “Intention to Buy” matters. How do those guys figure out so many ways to look at irrelevant data?
Review my previous article on this topic at: