What would Google look like if you took it offline and forced it to set up in physical space?
Google's tens of thousands of commodity Linux servers would need to be scattered around the globe so as to collect and then aggregate consumer interest. A lot like Path Intelligence, in other words.
Tim O'Reilly invested in Path Intelligence back in 2007. After lunching with Path Intelligence CEO Friday in London, I can see why. The idea is to set up receivers in shopping malls and other retail areas to collect mobile data and analyze consumer behavior to help optimize lease rates, retail setups, etc. It's a big idea. Huge, even.
But it's what comes after this initial phase that has my brain hopping.
Path Intelligence's initial phase could prove somewhat costly, as it seems to require a sales force and hardware development/assembly to gain a critical mass of property owners to adopt its service. (The software,, is free, which helps to keep costs down.)
But once Path Intelligence has a reasonable amount of distribution, the network effect of its business is absolute manna from heaven.
Just like Google.
Importantly, Path Intelligence is highly useful to retailers and others that want to optimize physical space (for retail, conferences/exhibitions, or whatever), even if the company never acquires another customer. If I'm an asset manager and want to know the most highly trafficked location in my shopping malls (e.g., so that I know where to charge the highest rent), I can easily get that data from Path Intelligence's service.
It dawned on me, however, that once a critical mass is attained, Path Intelligence can start selling subscriptions for the data to marketers everywhere, and could even start offering targeted advertisements to consumers while they're shopping. If I'm Ralph Lauren, I want to know that customers at Nordstrom who stop at the Faconnable display never make it to my displays, and then negotiate with Nordstrom to change my position.
And if I'm shopping at The Gap, I'd love to get an SMS ad that suggests I head to H&M for $10 off (perhaps sparked by data in the Path Intelligence system suggesting that a significant percentage of shoppers who go to The Gap later move on to H&M). It's good for me, and good for the retailer.
If this sounds like a privacy nightmare, Path Intelligence isn't blase about the issue. All information collected is anonymous and Oliver told me the company refuses business with organizations that may want to take a Big Brother approach to the data.
Google's advertising machine works because users constantly tell Google what they want. Path Intelligence, in physical space, is much the same.
It's a Very Big Idea, and very similar to Google. At scale, Google is unbelievably powerful because of the data that powers it. At scale, Path Intelligence could be the same.
Importantly, however, Google is useful to companies that simply want to search their internal documents or Web sites, just as Google Desktop is useful to individuals who want to unlock the clutter on their machines. Path Intelligence is immediately useful to retailers and property managers, but it becomes even more powerful once it has a sufficient body of data coursing through its servers.
Path Intelligence has just started raising its next round of funding. It's still in its early stages, but that's what makes it so exciting.
Follow me on Twitter @mjasay.