Until today Alexa mainly used it’s own toolbars’ data. Because its strong distribution in the group of techheads and webmasters, the compiled data has not been very representative in many cases. Alexa already tried to fix that by serious number crunching, but nevertheless sites with a tech affine readership had more “optimistic” charts, while sites with a broader focus came off badly.

Alexa data reloaded

Today Alexa inserted new data into its system. According to Alexa it is based on a wider range of data sources, of which one is the normalized global representative panel. Also the calculation method of the trend chart was optimized as well as the geographical distribution statistics.

At the moment the data only reaches back 9 months in time – Alexa already announced that the historical traffic statistics will be delivered in addition.

More info: Alexas publication of the new system

Article is also available in German: Alexa: verbesserter Algorithmus & neue Datenquellen

How do you like the new Alexa?

Maybe he could have done some things another way, to increase his chances of selling all or at least some more available ad slots? I think so, here are some ideas he could have considered:

1. When the advertisers were the ones to enter the draw (instead of the “normal” users clicking the ads), chances for selling more ads would have increased. There would have been fewer clicks onto the ads, but the advertisers should know that this is only junk traffic anyway. Even more ads would have been sold, when an advertiser could get a higher chance of winning the price through buying more ads.
2. $1 million may seem like a lot of money, but the chance that you are the one to get it is near zero. Splitting the money would have probably been a good idea (for instance: 1x$500.000, 2x$100.000, 4×50.000, 5x$20.000 etc.)
3. Lowering his own share could have also increased the chance of creating much more buzz. The student with “no money at all” who is selling pixels to afford his degree studies was a story that’s sticky but a millionaire trying to do another million is not.

As the sum to be given to the winner is $153.000 I reckon he must have earned the same amount (deducting server cost and managing), as every pixel was sold for $2. That is not too bad for a near copycat of his own idea and it’s even more he expected to gain through the original idea in the first place (he needed $37.000 to cover cost for 3 years at university).

Does anyone know something about Alex’ eBook about the hole story which he anounced to be writing in May 2006? It doesn’t look like it will be published anytime soon, isn’t it?

What do you think? Did Alex do the right thing, or could he have even doubled or tripled his income using another strategy? Let me know your opinion.

Update: it’s official now: the winner of the $153.000 draw is Mr K. Moguche from Kenya, as you can read on pixelotto.com.