Location Tracking and Heat Maps
Citysense by Sense Networks launched two days ago. Since the cat’s pretty much out of the bag, I might as well talk about this. Quote:
What if you could look at your cell phone and see a heat map of where everybody in the city was at that very moment? The more people at any given location, the redder it would appear on the map. That’s what Citysense does. It is a mobile application that is supposed to help you figure out where the hottest clubs and night spots are so you can go there (or avoid them, depending on your preference).
The reason why I find this so curious (and why this post has any mention of cats escaping bags) is that this is an idea Henry and I were strongly considering about two months ago. We wanted to use heat maps of cell phone activity to plot the hottest places. Since it would be done in near real-time, the theory was that it could help people find popular nightlife or help businesses find ideal retail locations.
We were so intrigued by the idea that we used it on our application for Y-Combinator’s summer round of funding. I even spent two weeks making a heat map demo that could render thousands of data points.
What happened? Y-Combinator didn’t invite us for an interview. We got a form letter back saying the equivalent of “better luck next time.” It was no surprise; they got enough applications to bring down their server right before the deadline. Our application probably just got lost in the shuffle.
After the Y-Combinator application was over and we got our response, we promptly dropped the concept. There were some problems with the idea (which I’ll get into), and we had other, related ideas to investigate. It’s funny how when you run your own company and can do anything, a lot of ideas come and go.
Now it’s two months later, and Sense Networks, maker of Citysense, is rolling out location tracking analytics and heat maps in a big way. They have an impressive set of executives, too. Of interest to me are the two guys with ties to the MIT Media Lab, Alex Pentland and Tony Jebara. I’ve probably seen some of their work before, especially Pentland and his Reality Mining.
In a way, this launch confirms that we were on to something, and it makes me wonder why we dropped the idea. However, the idea did have some drawbacks which we realized at the time.
The largest problem was data. How do you get enough data to actually have realistic and useful statistics? You could start from the ground up with each individual cell phone user, asking them to install an application that tracks them. From the user’s perspective, there are two main problems with this method:
- Any rational person will have huge concerns about the privacy implications of installing the tracking application on their cell phone.
- The user needs to perceive some benefit from the application, otherwise they won’t want it.
An application like Citysense is a great way to overcome these two problems. It aggregates anonymous data, and it’s branded as a fun way to find active nightlife. This was the same solution we came up with to solve the need for data.
That said, I’ve ignored the elephant in the room. No one will want to use the nightlife application if no one else is using it. For the first few users, there will be nothing to look at. You need thousands of people for the data to be any good.
How does one overcome this? Well there are other sources of data. You could go to cell phone companies and ask for access to their data. That would surely get you lots of information, but I’m sure the cost would be astronomical. Plus, cell phone companies aren’t know for being open with their data.
That’s where Sense Networks has an edge. They have prominent executives, namely guys like Pentland, who have long standing research relationships with major cellular companies. For us, location tracking and heat maps was an idea we had two months ago. For Pentland, it’s been a core focus of his research for years, and this is at MIT, a place known for having great connections.
Taking all of this into account, we are still interested in this area, and the launch of Sense Networks won’t stop us. There are massive opportunities waiting in mobile devices, and we are just beginning to see them bloom.
Aggregating, analyzing, and acting on the behavioral data that mobile devices generate will be one particularly hot area. It will enable contextual applications, and it will change the game similarly to how Google changed online advertising with contextual ads. I won’t say exactly what we’re doing yet, but we have a pipeline of products lined up to get involved in this particular opportunity.