16 Comments

Summary:

Local recommendations are becoming a key feature for location-based apps as they look to provide useful suggestions for users. Recommendations represent one way location services are moving beyond check-in, helping them transition from broadcasting a user’s current location to assisting them in finding their next one.

photo (2)

For location startups, the check-in is like wearing a shirt in a restaurant, without it, you have no service. Soon coupons or other deals will join the check-in as the sort of commodity offering that every location-based service must offer. But instead of a race to add more commodity-level services, there’s a place for these startups to differentiate — local recommendations. We’re forcing the web onto smaller devices we want to take with us and access over relatively small pipes. The end result of that behavioral shift is the emergence of the hyper-personalized web. A company that can deliver the best version of the hyper-personalized web on our handset (and that needs to include our location) and help users interact with the world around them will win. So how does the competition stack up for this next big battle for location (and even the web)?

There are different ways of implementing local recommendations. Some services are tapping directly into a user’s travel history to create personal recommendations, while others are asking for input from users about their interests or previous ratings for places. Some just enable notes left by a user for a specific location, which can serve as mini-recommendations. Here’s a run down of how many of the players are pursuing local recommendations.

Foursquare. Foursquare hasn’t instituted a recommendation engine, although CEO Denis Crowley announced in September that the company is working on one and a new job listing for a data scientist highlights those plans. The idea would be to take a user’s check-in history and create a list of suggested places that might suit their tastes. Crowley is also looking at enabling pop-up recommendations for users when they pass a place that a friend frequently visits. This would tap into Foursquare’s vast data from its 5 million users and, provided the algorithm is good, would give users an easy and passive way to know what else to check out.

Loopt. Location pioneer Loopt just updated its mobile apps with background notification updates. But another important improvement is the ability for a user to see places where their friends like to visit nearby. This doesn’t include full-blown recommendations, but leverages the check-ins of friends to show places a user might want to check out. Like Foursquare’s approach, this would be more passively derived but could be helpful in showing people places that they might have missed.

Where. Where also updated its app for iPhone, Android and RIM earlier this month with a new local discovery engine. Users looking for nearby “best bets” are asked to pick five of their favorite kinds of places for food, drink or entertainment. From there, the app is able to deliver local suggestions using semantic indexing. Where has also included a new guides feature that includes curated lists of places to go, and provides improved recommendations if a user connects to Facebook, allowing Where to tap their interests and likes.

Bizzy. This location start-up is trying to go down the Where route but with an even more aggressive questionnaire. Users who answer 20 questions about their favorite things online can get matched up to profiles of people with similar tastes. The Bizzy algorithm then provides a short list of places that match a person’s interests. This is reminiscent of Hunch’s taste profile questionnaire, but it felt a little cumbersome having to write in the answers of my favorite places, which is more tedious than Hunch’s A or B selections.

Google. Google has gotten into the game with its Hotpot feature, a local recommendation engine in its Places service. It takes the ratings a user and friends give local places, then uses that data to build recommendations for future places to check out. A user can also see friends’ comments for places when they search for something. Google obviously has a lot of potential in this area with its ability to tie Hotpot into Android’s Maps app. But it’s dependent on a user and friends’ ratings, which can be less stellar if the user or friends aren’t into reviewing places.

Gowalla. Gowalla doesn’t have a sophisticated recommendation engine. Its latest update was aimed more at tying into Foursquare and other services. But Gowalla has improved its Highlights feature so it’s accessible from Spots pages from the iPhone. Gowalla allows users to also view tips from Foursquare for a location. They’ve also updated Gowalla’s notes feature so a note can be left for a specific person at a location, giving tips on what to order or where to go nearby. Gowalla’s approach is more low-key, but it also shows another way to incorporate recommendations into a location app.

Others players to watch include Facebook, which is sitting on a mountain of data about a user’s likes and could become a major recommendation engine for all kinds of products and places. Hunch has also launched a feature called Hunch Local, which mixes data from Twitter and Facebook along with its users’ answers to help suggest places to go. And Bundle, a Citibank spinoff site, has also announced a restaurant recommender for NYC and Los Angeles.

Again, these are still early days. But with so much data available and companies under pressure to compete and become profitable, it only makes sense to see more of them turn to local recommendations to stand out. This will become quite competitive, but what will separate the best companies from the pack will be the underlying data and algorithms they rely on. That will allow them to be relevant and helpful in a world swimming with location data.

Related content from GigaOM Pro (sub req’d):

  1. Agreed! This is the very reason why we built a recommendation engine into TriOut ( http://towld.com/app ) from the ground up. Even though TriOut started off as a hyper-local location-based service, the information from recommending places that your friends checked into, rated and uploaded content from was valuable to our users.

    It will be interesting to see when or if the other LBA add recommendation engines to their platforms. I believe Whrrl has already implemented a recommendation engine into their app.

    Share
    1. When I saw this post I knew I’d see a comment by Mr. Sutton !

      Recommendations are the way to go. The fun and “game” of the mere check-in has played its course due, in small part, to no real incentive for checking into a location.

      Give me a tangible recommendation from my circle of influence and that makes go-location apps viable.

      Share
  2. And of course flook is all about recommendations and discovery – we always believed in the richness of that rather than a simple mechanism like check-in.

    Share
  3. [...] are …HOW TO: Create a Location-based Marketing CampaignPenn OlsonFoursquare ReviewPC WorldLBS Startups Should Forget Check-Ins and Just Tell Me Where To GoGigaOmThe Press Association -Independent -Huffington Postall 25 news [...]

    Share
  4. Whew! That’s a lot of services trying to give you local recommendations.

    Imagine if you could consolidate all this fragmented data you have across multiple networks such as foursquare, facebook, twitter, hunch, etc into one place and get useful recommendations based on ALL this location and preference data.
    That’s exactly what we’re trying to do at SpotOn.

    You can pre-register at http://www.getSpotOn.com with your foursquare or facebook account.

    Share
  5. We are taking a different approach at Bundle.com where we are using anonymous credit card transaction spend information to see what people are doing. We have a database of about 20 million people and have built a recommender for restaurants. So we can do things like assign a loyalty score to a restaurant based on how often people go back. The downside is that this is not realtime but the depth and validity of information makes it very interesting.

    Share
  6. [...] actions on the ad network but they were also able to point the company toward a new business: a local discovery engine for consumers that recommends places to go. That’s the value of data scientists said [...]

    Share
  7. can’t agree more! I think social recommendation should be a central pierce of LBS, like what facebook does.

    our innovative LBS (iphone, android, website) has a built-in support for social recommendation for both information and place. please see my post for more details
    “Opportunities and Problems of Foursquare-like LBS” http://nearbyfeed.wordpress.com/2010/12/09/opportunities-and-problems-of-foursquare-like-lbs/

    Share
  8. [...] services can break through beyond the check-in. Some are offering more deals, while many are turning to local recommendations. Priebatsch said SCVNGR isn’t looking at those two options because he feels like they are [...]

    Share
  9. [...] OM: From an average person’s standpoint, location is still kind of geeky and a lot of work. [...]

    Share
  10. [...] why it makes sense for services like Foursquare and other location companies to focus on discounts, recommendations and other practical [...]

    Share

Comments have been disabled for this post