Mario Sundar's Speakeasy

Spero Ventures. Early LinkedIn, Twitter. These are my thoughts on tech, brand, marketing and community.

Is Facebook’s Graph Search a Giant Killer?

Will Facebook’s “Graph Search” be a threat to Google, LinkedIn, Yelp, or Foursquare asks a question on Quora?


No, No, No and Definitely Not. Yet.

The key is expertise.

Beneath the obvious user delight, Facebook is betting a lot on Graph Search’s core ability to connect people with what they’re looking for accurately and immediately. And obviously as the middle man, they stand to gain. Fair enough.

But will Facebook’s imminent functionality be a threat to well established vertical searches like Google, Yelp, LinkedIn and Foursquare?

All of the four kinds of search you can do today: Photos, People, Places and Interests, bear commercial implication. But the most immediate remain People and Places, which as bloggers speculate may pose a threat to Yelp, Foursquare, Google (Places) and LinkedIn (People). So, let’s take simple examples and compare Facebook Search with the other four searches.

Facebook vs. Yelp

I started with a simple search for “bars,” something I presume will be a common search on any local product. Here’s what I got with Facebook. For starters, along with actual bars it also pulled up law and bar associations or offices which was a bit odd.

Photo Jan 19, 6 50 08 PM

Now try the same with Yelp and you see how right away, they try to segment that query into the different types of bars you’re potentially searching for.

Photo Jan 19, 6 50 33 PM

Once you get a set of results, Yelp then allows you (and this is the most useful feature on yelp currently) to convenience sort by “rating,” “proximity,” “price,” “open now,” or even better by neighborhoods.

Photo Jan 19, 6 50 58 PM

I’ve gotta tell you; if you go out often, this filter is magical. But again, the filter is by utilitarian ratings by foodies and not by friends around you. More on that in just a second.

But before we leave Yelp, the third most useful feature on Yelp is their surfacing key elements of the review. So you’re at a restaurant and you’re wondering what’s the best thing on the menu. In days past, you’d have had to ask the person serving you but now you can rely on “the wisdom of an expert crowd” what’s the best food here and it works. Like magic.

Photo Jan 19, 6 51 23 PM

Facebook vs. Foursquare

Back to the topic of friends which is Facebook’s biggest competitive advantage. If you do wanna take into account which restaurants your friends are frequenting (ignoring the fact that expertise is the key), then try Foursquare.

The first thing you’ll notice yet again is the structured data (categories like Bar, Sports Bar, Salon) right up front (similar to Yelp) that Foursquare now provides you; though not as in depth as Yelp, can still be a tad useful.

Screen Shot 2013-01-19 at 7.10.49 PM

Digging deeper through the results, you’re gonna find them sorted by Foursquare’s own proprietary “Zagat number” that they conjure based on multiple data points.

Foursquare comes up with its score by looking at tips left by users, likes, dislikes, popularity, check-ins and it also weights signals more heavily for local experts.

They also show you a self-selecting group of folks who you know. Chances are most of these folks are more prone to bar hop than your other friends. But still Yelp really nails it with their community that they have nurtured for many many years who continue to write meaningful reviews that makes a world of difference when it comes to local search.

Screen Shot 2013-01-19 at 8.36.37 PM

Facebook vs. Google Local

While on the topic of a Zagat number, Google recently bought restaurant ratings site Zagat which now powers their Google Local ratings.  Zagat which originally started off compiling restaurant ratings of the Zagat’s friends, does something very similar to Yelp and the model here is yet again – expertise.

Photo Jan 19, 7 22 43 PM

Facebook vs. LinkedIn

Shifting gears to people search, Facebook’s people search is three years after LinkedIn launched its faceted people search. I know because I helped launch it at TechCrunch Disrupt where product manager Esteban Kozak demoed it right before CEO Jeff Weiner went on stage. (Disclosure: I no longer work at LinkedIn and don’t own any stock either) My mind was blown when I first saw what we could do with faceted search on LinkedIn both from a user experience perspective and I’m sure recruiters have found even more value from it.

Take a look at this demo video we shot in 2009 that shows you the plethora of signals a site like LinkedIn uses to hone in on the right professionals in a search. Easier said than done, and much like with Yelp, these signals have been gathered over many many years and such a search isn’t something you can turn on willy-nilly.

In all four instances the quality of Facebook’s search is insipid today compared to the robust community based expertise that the four sites have either built or bought .

The key is expertise. 

Now granted there are many things Facebook could do to build or buy their way into each of these verticals but the key point is that strength in local search across People and Places is not “friend” related, but rather “expertise” dependent and it takes years to build that. And frankly, I’d go with the critical reviews from experts in these fields and that’s an area that Yelp, Foursquare, Google and LinkedIn have Facebook beat.

Filed under: Facebook, Google+, Linkedin, LinkedIn Features, Local Search, Location, , , , , ,

Why Google Circles is a Giant Fail

Google+ (Google’s new social offering) seeks to differentiate itself from Facebook with three key features: Google Circles (their tentpole feature, seemingly a retelling of Facebook privacy lists), Google Sparks (seriously useless) and Google Hangouts (which was my first positive reaction to the features Google+ launched with.

Let’s focus on Google Circles for a second, since I read a couple of posts today: one, by Kevin Cheng – product guy at Twitter, who shares my thoughts on why friend lists don’t work and another from Fred Wilson on the potential evolution of social suggestions by Google, and therein lies the rub…

So, let’s start with my Google Circles thoughts (originally published on Quora)

“A friend list by any other name…”

Call it Google+ Circles, but it’s still just a far easier-to-setup-and-use friend list. Facebook allows you to create different friend lists today and even prompts you to bucket your friends when you connect — but how many users use that.

Unfortunately, this ain’t a sustainable solution.

Here’s why: my friend lists are constantly evolving and are extremely nuanced and (no way) am I gonna continually update these lists over time. For e.g. I’ve created a list for Colleagues. Guess what? Colleagues move and so do you (from job to job), and once somebody leaves that Circle — that privacy list is useless — cos you’re now gonna share something that they shouldn’t be seeing. So, you’ve a broken list…

Yes, Google Circles may be good spin but it’s futile. Kevin echoes similar thoughts:

Thus, maintaining digital groups has two problems. First, you don’t know when to move someone from one group to another because transitions happen gradually. Second, it’s simply a lot of effort to maintain. How often would you update the entire list? And if it’s not updated, how useful are the groupings, really?

He then takes it one step further by wondering out loud if it’d be possible to automate these groups? Here’s Fred Wilson elaborating on this topic:

This is an oppportunity to use machines. And Google is doing this with Google+. The recommendations on who to add to what circles are amazing. So why make me do the drag and drop thing other than it is fun and cool to do that on a computer?

If Google+ knows who my music friends are then just suggest “music friends” when I hit the share button and send it on. Do I care if it goes to a few people who aren’t actually my music friends? No I do not. Do I care if a few of my music friends don’t get it? Yes, but then I can add them explicitly. I trust Google to do a fine job of this for me. They’ve proven themselves worthy of the job so many times in my relationship with them over the years. I trust that they can build algorithms like this as well or better than any other company out there.

Google doesn’t get social, yet.

I beg to differ. In my personal experience, I haven’t seen Google’s expertise at crafting good social recommendations, yet. And, here’s my take. As I’ve said from the get-go, Google+ seems less like an innovative shift or evolution in the social networking construct and I’m still not convinced that they get social or community.

As I’ve argued earlier, sites like Facebook, LinkedIn, Quora, Tumblr get social and are building their communities from the scratch. Google seems to be applying a sliver of social across all their existing products and I posit that they hope to create an identity ecosystem that ties it all together. Here’s an interesting theory from Vincent Wong that positions Google (and Google+, right) against Microsoft and less as a social play against Facebook.

Google’s Implicit Graph

But, I digress. Speaking of social recommendations: what does Google know today? Who do you email, from whom do you receive spam, whose emails do you ignore and what do you search for. A huge win for them in the past. But, an email graph plus search doesn’t a social graph make. That’s why I’ve wondered what sense it made to integrate Google+ Like buttons onto search.  Granted, Google+ is their first sorta serious attempt at social (Buzz, Wave, RIP), but frankly I haven’t had one pure social interaction with friends or family on Google+ yet. I’ve over 3000 followers (reminds me of Friendfeed and Twitter) but it’s definitely not the social graph we’re seeing here. It’s random comments from marketers, bloggers and Googlers seeking focus groups on Hangouts. I kid you not.

You know who could create the implicit social graph: Facebook can.

From day one, they’ve created a truly social universe for three quarters of a billion people and they probably possess a slew of real social gestures (whose walls do you visit, whose photos do you tag, whose tags of your photo do you untag, whose photos do you comment in, what group messages are you being sent today, whose parties are you invited to, which of those parties do you actually go to, I could go on… and then there’s the mobile side of it, places, etc.)

Now, I’m sure my experience may not be your Google+ experience (if yours is different please leave a comment or @mariosundar me). But, to summarize, I believe Facebook has a real good shot at creating the implicit groupings that Fred and Kevin talk about and I look forward to that.

And, btw, you know who can suggest music friends. Wait till Spotify and Facebook unleash their plans for world domination and then musical social recommendations won’t be too far fetched.

Disagree? Sure. Leave a comment or let’s chat on Twitter.

p.s. All said and done, I wouldn’t write off Google. They are definitely in this for the long haul, are taking a ton of feedback from users and who knows what their next play will be. Rest assured, you’ll hear about it here.

Filed under: Google+, , , , ,

Why Google will always remain Spock. Never Kirk.

The past few days have witnessed a barrage of non-stop Google Plus nonsense, with marketers vying with one another to carve out their territory on Google+ with the fond hope that it’ll be the next Twitter. In the meanwhile, I’ve not had one meaningful conversation on the platform with nearly 721 followers and I don’t know of any who have.

So, what gives?

Google+ 0 Friends

To get sticky with it: You always start with the community.

Let me share with you a tale of two other social sites that have increasingly become my daily go-to sites: Quora and Tumblr. Those who follow me have probably seen my tweets from either of these sites and the reason is, when I’m there I feel like home. In much the same way as I do on Facebook, which has my real friends and family.

Facebook started with the college community, built that flawlessly across the country, and then finally expanded outside of that circle that they had so masterfully cornered. This was probably what helped them break the monopoly of MySpace, whose ignominious ending we all witnessed this past week.

A tale of two useful social sites: Quora and Tumblr

Likewise, the kinship with my peers on Quora and Tumblr took months to form. On Quora we share a common interest in learning and several common topics that the site is carefully curating over time (like a good librarian who can direct you towards a book that you should read). Tumblr, likewise has a group of artful types who share quotes, pictures and videos (yet again, on topics I dig).

And, on both sites I find good search functionality that lets me pull in updates on these topics I love. Note: I wish both would automatically pull in my Facebook interests since they’re providing a high-quality stream of content on those topics that even Facebook cannot generate. Take that Google+ Sparks.

Now, I probably wouldn’t have published this if I’d not seen this morning’s top post on Techmeme from Paul Allen on Google+ that proclaims:

 Google+ is Growing Like Crazy. Report Coming Monday. Probably More than 4.5 Million Users Already

To which I say: So what? Actually, hang on, Business Insider says it better:

In fact, two days after Buzz went live, Google posted a blog entry bragging that “tens of millions” of people had checked it out, and created more than 9 million posts and comments.

At some point, interest died.

So far Google+ is filled with Googlers, reporters, and tech enthusiasts. They’re posting a lot, enjoying the Hangouts feature, and driving traffic to tech news sites.

But it’s still way too early to know whether Google+ will get any traction with mainstream users — the 750 million people who are on Facebook today.

Personally, despite having hundreds of followers on Google+ nothing of interest has happened on the site in my purview. Yes, I see my good old blogger friends asking questions they used to ask on Twitter, I’ve seen some cool hangouts with random people that Ben and others started, and the curiosity factor over which “interesting stranger” (as BI called it) is on G+ today. 


Google just doesn’t seem to get social. While the screenshot above (Googlers with 0 Friends) may be a great metaphor, as I’ve argued from the beginning, the Friendfeed cult model (that G+ mimics) just doesn’t work at building sustainable social communities, since it confuses the personal and public spheres. Granted it may scale faster as you’re gonna see soon (millions of users real fast), but will it stick?

Here’s a blog post from George Siemens that suggests why the friend forming algorithm of G+ is messed up:

While power laws (Pareto’s Principle) may exist in many areas of our lives – banking, TV watching habits, book purchases – they are surprisingly absent at a personal level. Yes, I likely respond to a small cluster of blogs and tweets that I encounter. But my personal networks – family and friends – don’t seem to have the power law structure of my public identity. For example, I move fairly fluidly between my personal networks. Facebook gets this. I’ve had very few “way out there” friend suggestions on Facebook.

G+, on the other hand, has been busy trying to make kings of a few: Robert Scoble, Mike Arrington, Loic Le Muer, Mark Zuckerberg, and so on. (Techcrunch addresses this issue as well.) I have precisely zero interest in those people. Nothing in my email history indicates that I would like to connect with them. Google’s algorithm is whacked on how it recommends friends: it is recommending them based on power laws (who is most popular) not on my personal interests. This is a fundamental and significant misunderstanding of social networks. Network properties are different at a personal and social level than they are in public spaces.

Welcome to the Friendfeed conundrum that conflates public and personal spaces. Even, the Pavlovian model of notifications is broken (and frankly useless) in this world, since now the red notification isn’t bringing in the reward that a Facebook notification does and is diminishing its effectiveness.

It’ll be interesting to see how Google+ evolves over time (cos they’ve really invested a ton of resources and are betting their future on it), but in its current avatar I don’t see how it can draw people away from Facebook.

Come back tomorrow for my post on Zuckerberg’s presentation style. This one’s a doozy. Bookmark my blog or subscribe to it.

Related posts you may find useful to form your own opinion:

  1. Follow the Quora topic on Google+
  2. Yishan Wong’s Quora answers (most of the recent ones are on Google+ and social)
  3. Ross Mayfield building on my original post re: different social networking models
  4. George Siemens post on Google+’s fundamental misunderstanding of networks
  5. Rocky Agrawal’s Solving the Scoble problem in Social Networks on TechCrunch (I’d say this is more of a G+ problem)

Sorry, Google+. For similar thoughts, follow me on Twitter.

Filed under: Google+, Quora, Tumblr

The Return of Friendfeed (as Google+)

I recently shared my thoughts on Google+ (Google’s recent foray into social networking — let me know if you need an invite — after the public failure of Google Wave and Buzz). They’ve got to get it right this time (and frankly I think they nailed some of the subtleties that they didn’t in their past avatars). That said, there’s just something about Google+ that doesn’t seem right and — that’s got to do with its relationship model.

Check out the rest of my thoughts on Google+ on Quora

Google Plus is a curious amalgam of Facebook and Twitter but more interestingly this is the same model that Friendfeed pioneered (with far slicker tools: “like” and “real-time feed” anyone).

Google+ is basically the 2nd coming of Friendfeed and therein lies the rub…

Information networks vs. social networks

Why is that a problem you ask. Let me first explain the two different types of social networking models. Traditional social networks (like Facebook, LinkedIn, etc.) have a symmetric connection model — mutual connections mark the relationship. Twitter on the other hand (an information network if you will), followed the subscription model where you follow users (much like Ev and Biz’s first hit –, which Google later bought). To the best of my knowledge, here’s the best description of the two models — hat tip to Joshua Porter (Bokardo), who did a terrific piece explaining these two social networking models. Highly recommended reading.

[Update]: Ben Parr just tweeted with his more recent piece, on Information networks. Here it is. Extrapolating, it’s basically any community that’s based on the information ties that you have (Twitter, Quora is a great example / follower model) vs. ones that are predicated on social ties (Facebook, LinkedIn / mutual connections).

Now, Friendfeed (later bought by Facebook – oh, what an intricate social web we weave), came out with a hybrid model which allows you to have both friends (subscribers) and followers. Oddly enough followers could pop up into your conversations as well. So, rather than being the best of both worlds, what you end up with is the worst of both worlds. Initially, there’s an incentive to build your REAL social network (a la Facebook) that Google+ is trying to foster with Circles, but at the same time they pollute that atmosphere with the follower model, where people you don’t know jump in with comments that you don’t feel like responding to.

That was the problem Friendfeed faced and that’ll be the problem that Google+ will inevitably encounter.

Secondly, Circles or Friend-lists are not scalable (though Google+ has perfected the art of persistent engagement to get users to bucket them – nicely done). What this means is that Google+ will gravitate toward the asymmetric or hybrid model (as it already has, wherein your stream will be sprinkled with random comments from people you don’t know).

What’s Google+’s future?

As I said in my Quora post, Facebook has nothing to fear from Google+. They both operate under completely different models. While Facebook is focused on building real relationships and has assiduously built an environment that reminds me of “Cheers” (see below), Google+ is slowly morphing into Friendfeed.

At the end of the day, I just wanted to be someplace…

…where everybody knows your name,
and they’re always glad you came.

you wanna be where everybody knows
your name.
That is Home. That is Facebook. (For me)

Now, granted the asymmetric model gains traction and followers fast, the question remains: is it sustainable? Time will tell. I won’t bet against Google given their enormous muscle and their ability to weave Plus into every Google interaction you have. But, I don’t think in its current state, Google+ will draw me away from my real home on Facebook.

What are your initial impressions of Google+? Leave a comment or @mariosundar.

Check out the rest of my thoughts on Google+ on Quora

Filed under: Google+, , ,