Homethinking is a NYC based startup in the real estate sector whose DNA lies in data--turning vast amounts of public data on housing sales, real estate agents and demographics on its head to empower buyers and sellers. For example, you can lookup the sales record of that agent who keeps pestering you to keep them in mind when you're ready to sell.
Recently, they released their Compare Neighborhoods feature--actually, I think of it like a translator. Moving from San Fran to New York? Check out what Big Apple neighborhoods are demographically and socioeconomically similar to the Misson or Noe Valley.
Here's a comparison for someone moving from Russian Hill into Brooklyn. Apparently, they should move to Carroll Gardens.
Here's what I like about it:
They took their existing dataset and thought about the real life problems that people need to solve. Interestingly enough, we're working on the same thing now at Path 101--trying to figure out how you could get into a particular job, industry, or company, given your current resume. Just getting a user from Point A to Point B is fantastic feature that more people need to think about because:
1) It simplifies the value proposition by making it goal oriented.
2) It informs you about the user--where are they now and where do they want to go is a lot more information than most people get out of a registration process.
3) It's functionality that could live outside of the registration-only part of the site, so you're proving your value before you ask for anything.
Here's what I think could be better... or at least improve the effectiveness.
Building something like this as a black box is really hard--and if you don't get it 100% right, you really don't give someone a lot of degrees of freedom or ways to improve it themselves. Maybe they want to emphasize a certain criteria more than others. Not only is that interesting data for you to record, but it keeps the user engaged and puts the ability to get their questions answered at the tips of their fingers.
So, I'd want to see this in a much more open format--where I could actually see the stats that are being compared.
The other thing I might do, and this could really get into some heavy, ugly, natural language/semantic analysis, is to pair up with Outside.in to parse out what people are talking about in each neighborhood. I have a feeling you'd come up with a lot of interesting keywords around restaurant types, ethnicities, and all sorts of interesting descriptions like "quaint", "hipster", "young", "families", etc... Either way, pulling Outside.in newsfeeds should be a must for any site that has anything to do with Neighborhoods.
A survey might also be interesting... like, "Do you have a family?" "Are there a lot of places to take your kids?" I'll bet you that you could do 20 questions that could tell a lot about a neighborhood that would improve the quality of these matches or give users more detail.
So check this out... what did you think? Does it make anyone want to move to Bay Ridge in Brooklyn?
Note: While I had totally meant to get around to writing about this feature release, I didn't really focus on it until Niki answered the call to participate in my Donor's Challenge. For a $200 donation, I agreed to review the Neighborhoods feature. Want in? E-mail me! charlie at path101 dot com