Today I came across an interesting blog entry about how Flickr used the meta data in the images of it’s database to publish a “Market Research” study on camera popularity (scroll down to the bottom of the Flicker page for the interesting stuff).
So Flickr mined their image database and pulled out the model info from the image meta data (automatically entered by almost all digital cameras, and apparently some phone cameras too). They then normalized this data to show the most popular cameras in the Flickr community.
They don’t call it a market research, because it’s self filtering based on the members of Flickr (which generally doesn’t include the professional photographers). But the potential does exist to use such data as a market research study. Image if you had a wider audience of photographers (like photo.net or an internet based stock agency like istockphoto or fotolibra) and did the same thing…then you would have something close to market data.
One key thing to keep in mind…don’t take off your critical thinking caps. A “market research” of this nature has huge potential for being flawed…all based on how the data is interpreted. As the saying goes…there’s lies, damn lies, and statistics. It would be easy to just total the number of photos taken from camera a and camera b and say which one is more popular. But that would be wrong. You need to take into consideration the human nature…one person may upload more than another and their camera would skew the results. The real point is transparency on how the numbers are calculated is really needed…
Maybe a more important question for us photographers is this: are you gonna leave your meta data intact when uploading?
When I upload images (which it usually only to my own photography website), I filter out all but the key meta data. If I uploaded images to any other websites, I strip out everything but my contact and copyright info (easy enough to do with a script).