I have compiled a large dataset of what people are selling on Craigslist. This is for my mobile classifieds startup (UPDATE: the app is no longer available). The small release is from Boston Craigslist only, and it’s only raw percentages. One interesting finding is how vehicle-related classifieds dominate Craigslist. Another is the huge swings among some categories, depending on what time of the day/week/year the sample is taken. Read the Invantory blog post to learn more.

I have other Craigslist data as well, including estimates of market size, average price, and data from other times and from other areas. I’ll probably release some of it in the weeks or months to come, but really want to focus on our new software platform for local classifieds.

One interesting fact about the dataset is the connection to my Extension School studies. My ALM thesis was based on a computer content analysis of a news database to discern various policy trends. For many months, I spent my nights and weekends sitting in front of a computer screen, copying and pasting results from the LexisNexis Academic to a gigantic Excel spreadsheet, and then taking sample news articles and subjecting them to a computer content analysis using a software program called Yoshikoder. The Craigslist data was compiled in a similar way. While I didn’t use Yoshikoder this time around, I did perform a manual coding process on the Craigslist Boston data which I used to create market sizing estimates. To learn more about my ALM thesis at the Extension School, google “Harvard Extended” and read the first entry.