Today’s release of the latest version of Foursquare has me super excited. It puts together all of the data and capabilities the team there has built over the last five years to provide the best local recommendations: based on lots of detailed data and knowing your tastes. Dennis's vision for Foursquare has always been to create the best local discovery engine – one that’s personalized just for you – by crowdsourcing bits of information from users all over the world. Today brings Foursquare a big step closer to that vision.

Unlike traditional review-based systems, Foursquare gathers lots of information on where people actually go and spend time and what they recommend. So instead of simply providing an average of a number of stars Foursquare computes a score. Because of the many signals that go into this score it is difficult to manipulate by any one individual (such as the venue’s owner or drive-by review). As another important innovation, Foursquare furthermore takes your specific tastes into account when making recommendations. When you start the new app for the first time it will prompt you with possible tastes (these are derived from your past usage but you can confirm and change them). Those tastes are then highlighted in the recommendations and you can change them over time.

With this new release there is also a different privacy model. Everything in the new Foursquare app is explicitly public and that includes a new asymmetric follower model. It means I can follow anyone to see the tips that they are leaving and more importantly have the recommendations I receive be influenced by experts that I have personally curated. You can go ahead and follow me – I am not yet an expert on anything although I am getting close on airports, American food and Chelsea.

This completely public model became possible by splitting out checkins (the privacy sensitive part) into their own app called Swarm. Swarm retains the symmetric friends model, which means that you are sharing your location only with people you have confirmed. Better yet, with Swarm you no longer even need to check in. The app knows where you are and shares that location with your confirmed friends (only at the neighborhood level – to provide a detailed location you check in). Again, this passive location sharing has become possible without being a drain on your battery through the company’s technology and accumulated data.

Both Swarm and Foursquare will continue to improve over the coming months. That’s not just because the team has a great roadmap and will also observe user behavior and feedback but also because the system constantly gets better as more people use it. Speaking of which, I am on the road today traveling up to Massachusetts and will shortly be using Foursquare to find a lunch spot.

Today’s release of the latest version of Foursquare has me super excited. It puts together all of the data and capabilities the team there has built over the last five years to provide the best local recommendations: based on lots of detailed data and knowing your tastes. Dennis's vision for Foursquare has always been to create the best local discovery engine – one that’s personalized just for you – by crowdsourcing bits of information from users all over the world. Today brings Foursquare a big step closer to that vision.

Unlike traditional review-based systems, Foursquare gathers lots of information on where people actually go and spend time and what they recommend. So instead of simply providing an average of a number of stars Foursquare computes a score. Because of the many signals that go into this score it is difficult to manipulate by any one individual (such as the venue’s owner or drive-by review). As another important innovation, Foursquare furthermore takes your specific tastes into account when making recommendations. When you start the new app for the first time it will prompt you with possible tastes (these are derived from your past usage but you can confirm and change them). Those tastes are then highlighted in the recommendations and you can change them over time.

With this new release there is also a different privacy model. Everything in the new Foursquare app is explicitly public and that includes a new asymmetric follower model. It means I can follow anyone to see the tips that they are leaving and more importantly have the recommendations I receive be influenced by experts that I have personally curated. You can go ahead and follow me – I am not yet an expert on anything although I am getting close on airports, American food and Chelsea.

This completely public model became possible by splitting out checkins (the privacy sensitive part) into their own app called Swarm. Swarm retains the symmetric friends model, which means that you are sharing your location only with people you have confirmed. Better yet, with Swarm you no longer even need to check in. The app knows where you are and shares that location with your confirmed friends (only at the neighborhood level – to provide a detailed location you check in). Again, this passive location sharing has become possible without being a drain on your battery through the company’s technology and accumulated data.

Both Swarm and Foursquare will continue to improve over the coming months. That’s not just because the team has a great roadmap and will also observe user behavior and feedback but also because the system constantly gets better as more people use it. Speaking of which, I am on the road today traveling up to Massachusetts and will shortly be using Foursquare to find a lunch spot.