reDesign

August 20, 2010

Facebook Places is at the beginning of a long road

Filed under: facebook, foursquare, geotagging, lbs, maps, mobile, twitter, wireless — Tags: — Rakesh Agrawal @ 8:39 am
Facebook Places on the iPhone

Facebook Places on the iPhone

Facebook’s much awaited Places product finally launched this week. It’s the first step toward bringing friend finding to the masses.

People have been using Facebook to do this for years; posting their location in freeform status updates that their friends can read and comment on. (e.g. “heading to Cambridge for dinner.”) By turning that freeform text into structured location data, Facebook can make that data more useful.

From an iPhone or HTML5-capable mobile device, you can check in to a place, such as a restaurant, bar, movie theater, airport. You can also leave a message with the check in. The check in is posted to your wall and may appears in friends’ news feeds. On the mobile side, you can see a list of your friends and where they’ve checked in. Clicking on a place will show you details of the place, including a map and who has checked in.

The initial release is fairly simple. In fact, it’s not that much more useful than the freeform status updates.

Facebook is entering a very crowded space with competitors such as foursquare, Gowalla, Loopt, Google Latitude, Whrrl and Twitter. Many of those products are much more robust. Facebook’s key advantage is the size of its social graph: within the past 24 hours, 18 of my friends have checked in.

There are many opportunities for improvement to Facebook Places:

  • Basic UI. Check ins are sorted by time, not distance. A friend checking in 2,000 miles away 2 minutes ago is less relevant than someone checking in 2 miles away 5 minutes ago. This problem is exacerbated by the fact that the city isn’t shown. Considering that many people use Facebook to keep track of friends all around the world, this is a significant issue. Foursquare has a separate bucket of “Friends in other cities.” Update: Facebook now has a separate grouping of nearby friends.
  • Map view. Often, visualizing your friends on a map is much easier than scanning a list. Foursquare already offers this.
  • Visiting friends. Out of town friends who are in town aren’t indicated. One of the big potential values of social friend finding is discovering when friends are in town. If a friend from far away is visiting, I’m more inclined to want to get together than someone who lives in town.
  • Pictures. There is no way to associate a picture with a check in. Given the difficulty in typing on mobile devices, often a picture gives a lot more information. These pictures could also be used to build a much more robust Place page.
  • Pushing location. Sending people your location via SMS is tedious. You have to address the message, type out where you are. If they don’t know where it is, they have to pull up a map or text you back for directions. With Places, it would be easy to push a notification to friends with where you are, complete with map. This could be sent as a push notification on iPhones or as an SMS with a URL for other phones.

As with most Facebook product launches, questions of privacy come up. In general, I think Facebook has done a good job with the default privacy settings on Places. You must explicitly check in; there is no background tracking.

Only your friends can see where you’ve checked in. Unfortunately, my social graph on Facebook wasn’t designed with location in mind. When I decided whether or not to accept friend requests on foursquare, I used a tighter filter than on Facebook. Now, I’ll have to go back through Facebook friends and create a list of who should have access to location. (See Post technology columnist Rob Pegoraro’s piece on how he classifies his friends.) Yes, old high school friends have been known to burgle homes based on Facebook updates. If that worries you, watch Rob’s video on how to adjust your privacy settings for Places.

The one big complaint I have with the privacy defaults is that your friends can check you into a location without your permission.

See also:

About these ads

August 19, 2010

Heading toward the Facebook recommendation engine

Filed under: facebook, lbs, local search, maps, yellow pages, yelp — Tags: — Rakesh Agrawal @ 9:13 am
Image representing Facebook as depicted in Cru...

Image via CrunchBase

There’s an interesting thread over at Mike Blumenthal’s blog on the effect of Facebook Places on the local reviews space.

My view is that reviews and updates will coexist, much as blogs and Twitter coexist. People who were less committed to reviews will migrate their activity to Facebook Places updates. But Places could lead toward the ultimate recommendation engine.

In the local space, there’s really only one review site that matters: Yelp. They’ve got a strong set of tools and an active and engaged community. New restaurants and bars, which are often of the most interest, will have a dozen reviews on Yelp a year before they even show up on many Yellow Pages sites.

There are three big challenges with Yelp:

  • It’s been too successful. Many restaurants have hundreds of reviews. Although Yelp provides great tools for analyzing the data, it can still feel overwhelming. It also discourages participation from more casual users. In the early days of Yelp, I was an active reviewer. That’s tapered off substantially — what’s the marginal benefit of me writing the 426th review of a place?
  • These aren’t my real friends. I don’t know how compatible their tastes are with mine. It also affects the propensity to write reviews. People are more likely to do something that helps their friends than something that helps a generic audience.
  • Skewed demographics. Yelp primarily caters to a young, urban demographic. If you’re a mom in the suburbs, its value is more limited.

Facebook Places lowers the bar to participation and ties it into real-life social networks. Instead of writing out a long review, a few clicks is all it takes. Combine that with Facebook’s large user base on mobile devices — its monthly uniques on mobile devices is 4x Yelp’s monthly uniques on the Web — and we’ll see a tsunami of local data. (For more on importance of massive amounts of data, watch Google’s Peter Norvig’s talk.)

While each blip may not be as rich as the data in Yelp, you could build a recommendation engine to infer a lot from that data.

If I see that a place I am considering visiting is regularly frequented by my friends with families, I can infer that it is good for kids. Positive reviews can be inferred by friends going back to a place regularly. There are some friends who I have negative taste relationships with. If I know that they’re regulars somewhere, I know not to go there. Facebook can also make recommendations based on places I’ve visited and the overlaps with places my friends have visited. Facebook also has real demographic information which could be used to tailor recommendations.

Status updates in the social network also prompt discussions. Even if the original poster doesn’t write a review, it may be followed up by “hey, I was thinking of going there. what did you think of it?” Facebook could also close the loop by prompting people to add star ratings, Like or add comments a few days after a check in.

When it comes to restaurant reviews and recommendations, most people are looking for “good enough”. While you could spend hours reading every Yelp review of several restaurants and possibly get a better answer, a recommendation based on your friends’ activity is probably nearly as good. Facebook has done really well with good enough; Facebook Photos dominates online photo sharing, despite many functional weaknesses when compared with flickr.

I built a prototype of this when I was at AOL Search and even with a few users in the system, it worked really well.

More on: Facebook, local search, Yelp

See also:

June 24, 2010

Checking in with foursquare at SFO

Filed under: audio, foursquare, geotagging, lbs, local search, maps, mobile, social networking, twitter — Tags: — Rakesh Agrawal @ 12:57 pm
SFO is a hotbed of foursquare activity

SFO is a hotbed of foursquare activity. Creative Commons image by Håkan Dahlström.

With the increasing use of mobile applications such as Yelp and foursquare, it’s becoming possible to pull ideas from thin air. Users of these apps can leave tips for others to find that are linked to a specific location.

In most places there aren’t enough tips yet to make filtering an issue. San Francisco International Airport, with more than 57,000 checkins on foursquare, is an exception. It offers a glimpse of what we can expect as these services become more popular. The airport is the perfect petri dish for tips: it serves a technically savvy audience and people often find themselves there with plenty of time on their hands.

The SFO tips page contains dozens of notes including places to eat, complaints, ground transportation, wifi and power availability. Mixed in to all of this are ads, other spam and random observations. Some examples:

have a corned beef sandwich at max’s if you’re flying southwest. the best! well, really good

When you enter short term parking do it as far to the right as you can (lvl 2) & then immediately head to lvl 1. There is always parking next to gate and that is the lvl that connects to the terminal

Free wifi at the Continental lounge in Terminal 1- be warned, it’s located outside Security

Smoking hot brunette woman at gate 20. Stop by and smile at her. She is so lovely!

Bart to Millbrae gets you within 1 block of an in n out burger. Great for 3+ hour layover!

Heading to wine country? Take a moment to stop by St. Supery in the heart of Napa on Hwy 29. Mention this to get a 2 for 1 tasting.

Sorting through the volume of tips can be overwhelming. As the volume increases, we’ll need ways to filter them. Among the ways to filter:

  • Timeliness. Some of the tips, such as wifi at the Continental lounge, are evergreen. Others, like the smoking hot brunette are very timely. Tipsters should be able to flag their tips to self destruct. As I wrote earlier, being able to identify tips by timeliness would allow for new applications, such as sharing rides. (“Anyone want to split a cab to Moscone?”)
  • Social network. Among the tips were tips from people I follow on Twitter, including Danny Sullivan and Adam Lasnik. Being able to surface these would increase relevance.
  • Ads vs. not ads. Sometimes people want ads, especially if it can save them money.
  • Keyword search.

Places like airports are especially complex because they’re really collections of places, sometimes with other groupings and physical restrictions. Being able to filter tips by terminal would also be useful. But then maybe that’s best left to GateGuru.

June 16, 2010

Now we’re going Places

Filed under: geotagging, gps, lbs, local search, maps, mobile, social networking, twitter — Tags: — Rakesh Agrawal @ 1:59 pm

I’ve been writing about Twitter and location since my first post about Twitter in 2007. This week, Twitter launched Places, which allows users to add their location to a tweet.

Here’s a screenshot from 2007:

Twitter location 2007

Embedding location in a tweet the hard way in 2007

and today:

Embedding location in a tweet in 2010

In 2007, I used a third-party application from Where to include my location. Clicking on that link would take you to a map on Where’s site showing the address. (The link in the original post no longer works.)

With the launch of Twitter Places, the search is done within the Web browser (and soon in Twitter’s mobile applications). You can select where you are from a list of nearby places. Clicking on the place name brings up the map above and the option to view tweets about that place.

Although the difference between the two may seem subtle, they are significant:

  • Because the place is metadata, it doesn’t count toward the 140 character limit.
  • Place names are human readable, unlike addresses and latitude/longitude. Knowing the name of a place makes it much easier to find than just a street address, especially in dense metropolitan areas.
  • Places are unique to a specific venue. Doing a pure location-based search would return tweets from surrounding businesses or businesses that have since disappeared.
  • Integration in to the main Twitter experience means broad exposure and eventual standardization of place identifiers. That has been a longstanding challenge in the local space.

Twitter’s geo APIs have been available for several months and third parties like bing have created interesting applications like Twitter Maps. With the availability of places across the Twitter platform, we can expect to see more interesting applications including both real-time applications (ride sharing and ticket exchanges) and historical (restaurant reviews, past events).

Once Twitter allows owners to claim their Place and associate it with a Twitter account, we could see official tweets of announcements and offers incorporated into a Place’s search results.

When pictures are tagged to a Place (instead of a lat/long), we’ll have the ability to visually browse a venue in Twitter.

May 11, 2010

Geo-enabled Twitter comes alive on Twitter Maps

Filed under: bing, geotagging, lbs, local search, maps, mobile, social networking, twitter — Rakesh Agrawal @ 8:56 am
Bing's Twitter Maps show you what's going on

Bing's Twitter Maps show you what's going on

I’ve been playing with Bing’s Twitter Maps lately and it’s one of the better implementations of Twitter’s geo APIs that were introduced last fall. It shows tweets within the last 7 days plotted on the map. Google Maps recently introduced a similar feature, but it seems to only show items that are fed through Google Buzz (including tweets that people have configured to send to Buzz).

Some future applications of geo-enabled Tweets:

  • Events. For last-minute party goers, a real time view of what’s going on around town, complete with pictures and real-time reactions.
  • Ticket scalping. Rather than walk around for blocks talking to scalpers about what they have, glance at a list of tickets posted. The information transparency would result in a higher price to sellers and a lower price to buyers than what scalpers typically offer. (In my experience at baseball games, scalpers usually ask at least 3x what they paid.)
  • Finding a place to go. When in new cities, it’s often hard to figure out where to go — what are the lively neighborhoods at night. By looking at a map of recent tweets, you could quickly discover where people are still awake.
  • Read reviews from friends. Geo-enabled tweets filtered by those you follow would provide socially relevant recommendations.
  • Offers from local businesses. These could be persistent or distressed inventory. Slow night? Tweet an offer to draw in customers.
  • News. Twitter has long been used for user-generated breaking news. With geo-enabled tweets, breaking news could be aggregated by location in addition to hashtags. The biggest stories could be identified by an increase of tweets from a location (versus normal) and retweet frequency. News from media outlets could also be plotted.
  • Construction and accident information. Avoid bottlenecks by seeing tweets from fellow drivers, DOTs and news sites.
  • Trip sharing. Find others at the airport headed your way, cutting costs and reducing pollution.

And, of course, there’s friend finding, which is the most talked about use of geo-enabled tweets.

So far, the percentage of tweets I see with geo information is tiny (>1% of those I follow). But as more and more geotagged data is put into Twitter, the key will be applications providing the right tools to filter all of that data. At a minimum, we’ll need the ability to filter by time of tweet, people we’re following, hashtag and application (e.g. foursquare).

Unfortunately, bing’s Twitter Maps doesn’t seem to be available where real-time information would be most useful — on mobile devices.

More on: geotagging, social networking, Twitter

August 5, 2009

Past, present and future of online maps

Filed under: bing, google, local search, maps, microsoft — Rakesh Agrawal @ 11:55 am

Business names and landmarks on Google MapsGoogle announced yesterday that it has added more detail on its maps, highlighting businesses and landmarks. They even solved the Albert Einstein Memorial problem that I wrote about last year.

Businesses and landmarks are important because they make maps more in line with the way people think, instead of the way that computers operate. This change also means that businesses won’t have to resort to painting their rooftops to be easily identifiable.

There are two big challenges with what Google is doing:

We’ve come a long way from the early days of the Web when maps consisted largely of roads and a clunky user interface. We’ve seen the addition of aerial imagery, building outlines, photos, public transit, Street View, neighborhoods, user-generated content and live traffic. Google has driven much of this innovation, although to be fair MapQuest had aerial imagery first and A9 had a version of street view early on.

There is still a lot of work to do to improve maps:

  • College and corporate campuses. Campuses such as Google’s and Microsoft’s buildings have numbers, but these aren’t shown on the map. If you were meeting someone, they’d probably tell you to go to “Building 43″. My friend Adam at Google keeps a custom Google map to show where his building is. (Oddly, Microsoft’s Bing maps show building numbers for the Microsoft campus, but don’t let you search for them.) The same thing applies for airport terminals.
  • Controlled-access facilities. Businesses in controlled-access facilities should be hidden by default — few people are going to park and go through security to eat at an airport restaurant. On the other hand, if I’m in the airport, I want to know what businesses are in my terminal.
  • Handling nonstandard locations. Databases are organized around cities and states in the United States. This works for most places, but is problematic in areas that don’t follow the convention like Hawaii or Las Vegas. Hawaiians talks about islands, but the local databases don’t know the concept of an island. This is made worse by the fact that the same town name is used on multiple islands — there’s a Waimea on Kauai and Hawaii and a Kailua on Oahu and Hawaii. Local constructs such as “North Shore” and “South Shore” aren’t understood either. Navigating using local search on my recent trips to Hawaii was error filled.
    In Vegas, hotels are a primary navigation construct and many of those hotels have more shops and restaurants than do a lot of American towns.
    Given how popular these destinations are, I’m surprised this problem hasn’t been solved.
  • Parking availability. In a big city it’s rare that you can drive up to your destination and park right in front; finding parking can easily add 15-20 minutes to your trip. Companies like Urban Mapping are already collecting this information. I had a book called Park It Here! that showed street parking restrictions for every block in Manhattan. I’d love to see that data online.

More on: google, maps, local search

September 10, 2008

You too can be Rand McNally

Filed under: city guides, flickr, geotagging, google, gps, iphone, lbs, local search, maps, satellite navigation — Rakesh Agrawal @ 9:01 pm

How long does it take to get from Pike Place Market to Mount Rainier National Park? According to Google Maps, it’s a 5 minute walk; less than 1/3 of a mile. Pretty easy, huh?

Mount Rainier looks an awful lot like an office tower to me.

Mount Rainier looks an awful lot like an office tower to me. That'll be one steep climb!

In reality it’s a 2 1/2 hour drive.

This error is the result of a fundamental challenge in how data is collected for online navigation and local search products: all of the major services use data that weren’t originally collected for navigation. The typical source is phone books listings, which were originally compiled by companies like InfoUSA to sell mailing lists. In this particular case, the national park likely has a business office in Downtown Seattle.

Einstein monumentThe data have a distinct bias: toward places with phone numbers. While you can get exact driving directions to a local ball bearing distributor, look for a park, landmark or trailhead and you’ll often get erroneous data or no results at all. Occasionally, you’ll luck out and there will be business names nearby that incorporate the place name.

Just for fun, try finding the Albert Einstein statue in Washington, DC using an online mapping service. (If you can’t find it, click on the image to see the answer.)

On Sunday, I was looking for Hippie Hill in Golden Gate Park. Yahoo! Maps struck out entirely. Mapquest delivered a bunch of unrelated results.

An answer lies in another Yahoo! property: flickr’s database of geotagged photos. Although flickr’s search tools aren’t optimized for local search, it’s content is a great data source. The first search result in Google Maps for Hippie Hill is spot on; it came from Google indexing flickr’s geodata. The commercial results on the same map (the red markers) are mostly garbage.

Flickr is just one tool. As more people adopt GPS-enabled phones like the iPhone, we’ll see more and more user generated data. Enabling mobile data collection is reason enough to build an iPhone app.

Developing countries have the potential to leapfrog the developed world in data quality. Google’s Map Maker allows anyone to populate data in countries without large commercial data providers. You can pull up a satellite view and trace roads. Zoom into buildings and add local businesses. Find a park and add in key features. Because the content is user generated, the places that people look for the most will be added first, as opposed to the local ball bearing distributor. Places with colloquial names will become findable.

User-generated content has its challenges, of course. In the initial stages, tools like Map Maker will attract the community organizers who are passionate about their neighborhoods. Once the data start getting even modest usage, spammers will attack. But many of the standard spam fighting techniques can be used to deter them.

OpenStreetMap and Wikimapia offer similar functionality. OpenStreetMap is focused on creating and editing roads; Wikimapia is focused on places of interest.

Google Map Maker’s user interface is more intuitive than OpenStreetMap and Wikimapia. Unfortunately, Google is being hypocritical in how it treats the data. While Google Maps combines data from other sources, such as indexing geodata from flickr,  it is keeping content generated by Map Maker users locked up. Users can’t even download the data they create.

More on: geotagging, local search, maps

August 27, 2008

The Russians are coming! The Russians are coming!

Filed under: geotagging, google, journalism, local search, maps, media, newspapers, web 2.0 — Rakesh Agrawal @ 10:52 pm

I’ve long advocated that news organizations geotag the news. But I’ve been skeptical of automated systems for doing this. Google News recently provided a terrific example of what can happen when you use entity extraction for such a task:

Where in the world is Georgia?

Where in the world is Georgia?

In this case, reported by Valleywag, Google is comically wrong. But even when Google is roughly right, the map is often there just for the sake of having a map. The location information is often not very precise or isn’t really relevant.

For example, this story about a Yankees game puts Yankee Stadium somewhere near City Hall. Stories about national issues are often datelined New York or Washington because the reporter happens to be sitting in one of those two cities.

For individual story pages, an inaccurate map isn’t the worst thing in the world. But when you plot many of these stories on a map, they become worthless. In Google Earth, you can get a layer that provides geotagged news from The New York Times. I’ve seen pointless geotagging such as a story titled “U.S. Moves Toward International Accounting Rule” geotagged as being in the “USA”. (Which Google Earth plots in Oklahoma.)

There are many cases where geocoding makes sense and provides users a real service:

  • Restaurant reviews
  • Crime stories
  • Event listings
  • Travel stories

In each of these cases, the location is a critical part of the story. The minimal extra effort involved in geotagging these stories would significantly increase their shelf life and usability.

August 20, 2008

How’s the weather out there?

Filed under: geotagging, google, maps, social networking, twitter, web 2.0 — Rakesh Agrawal @ 3:20 pm

One of the things that takes getting used to living in the Bay Area is the many microclimates. Temperatures and conditions can vary dramatically within a few miles.

I was deciding whether to head out to the Beach Chalet, a microbrewery and restaurant on the Pacific Ocean, on Sunday. They’ve got a great back yard with Adirondack chairs and live music. It’s a gorgeous place — on a sunny day. Unfortunately, it’s located in one of the foggiest parts of San Francisco.

Before trekking all the way to the other side of the city I wanted to know whether it was sunny there. Could social networking help?

I decided to post the question — to noone in particular — on Twitter:

Wonder if it is sunny there. My question posed on Twitter.

Within 10 minutes, I had my answer:

Its Not.

Of course, this particular experiment is likely only to work this well this quickly in the Bay Area, where it’s easier to find a Twitter user than a newspaper reader.

How else could I find the answer? Google now has geo-coded Webcams in Google Earth. Find where you want to be and check to see if there’s a camera. Click and you’ll get your answer. In the case of the Beach Chalet, it’s often this:

No, I didn't just paste in a white image. This is a view from a webcam near Ocean Beach.

No, I didn't take a picture of a white wall. This is a view from the Cliff House webcam near Ocean Beach.

June 3, 2008

Occasional reader – Saying no to Google, popcorn prices popping, economics and height, Pringles

Filed under: fun, google, maps, movies, privacy, random, reader, street view, weekly reader — Rakesh Agrawal @ 1:09 am
Older Posts »

Theme: Silver is the New Black. Blog at WordPress.com.

Follow

Get every new post delivered to your Inbox.

Join 103 other followers

%d bloggers like this: