Update: as of 20080724-15:00 EDT apple has nulled out the data, the experiment is over
Update 2: as of 20080726-12:00 EDT the experiment is back on. My 2nd draft can be seen in the same spot: http://andrew.io/3g-retail-stock. It now has zoomable graphs.
I live about an hour from the nearest apple store, and HATE standing in lines. So on Monday, I started to wonder what is the best way to predict iPhone 3G shipments without stalking the store, and could I perhaps predict accurately enough so that I could get one without standing in line.
Using Apples retail availability page, which is only available after 9pm, was out of the question. But I was curious how they were getting data to that page, and was there a way I could get the data, collect it and analyze it.
And it turns out there was.
Apple is using a json data feed to populate their own availability page. And the json data feed is updating live 24 hours a day. So Monday afternoon I started collecting it about every hour and on Tuesday I started some basic visualizations.
My first cut can be seen at http://andrew.io/3g-retail-stock.
I don’t have a iPhone 3G yet, but I plan to continue collecting the data, and adding visualizations to see if any detectable patterns emerge over time. But at a minimum this can be viewed as an hourly availability tracker with some history of availability.
I hope people like it.
But if you use it, I would recommend you go ahead and call the store before you show up, cause if they don’t have them I take no responsibility.
