Last month, Facebook unveiled a major update to its search service that let users find status updates, photos, links, and videos, instead of just entries for profiles and groups. The speculation here (and elsewhere) was that Facebook, which has introduced Twitter-like features throughout its site, was once again trying to one-up its rival, this time with a real-time search service that could also take the “pulse” of the web. But in a recent interview with paidContent at Facebook’s new headquarters in Palo Alto, Akhil Wable, who leads Facebook’s nine-person search-engineering team, said that the company did not have any ambitions to take on Twitter in search. He also talked about how Facebook search might change, how Facebook may monetize it, and whether anybody clicks on those Bing results that Facebook includes. Read on for details.
paidContent: What’s the vision for Facebook search?
Akhil Wable: The main purpose for what we launched was finding out what your friends are saying about a particular topic. You have the news feed where you can just get on there and see what everybody (you) care about is talking about, but if you want to get down to a specific topic we didn’t have the mechanism to do that. So the big focus this time (is) how do you make it possible for you to search your friends’ content.
We do have two sections — one is the friend section, one is what everybody is talking about. I think the everyone part is just kind of a cool addition. If you don’t have enough information from what your friends are saying about a particular topic we don’t want to just leave you with nothing.
Do you see the focus of Facebook search changing?
No, I actually don’t. When we look at where Facebook provides value it really is understanding the user, the things they care about, things that they are connected to. That, sort of, will always be the core focus.
I assume you don’t think you could ever take the place of a Google (NSDQ: GOOG) search then.
We usually don’t even look a year ahead or three years ahead, but what we’re focusing on right now is things like making it really easy for you to find the people you’re looking for. Especially new users they get on to the site, it’s really important that they create those connections that will make Facebook valuable for them.
When we start talking about web search out there and are we going to start competing with Google, I would say, no, that’s not the way we’re approaching search today. We don’t sort of want to become this independent entry point. It’s mostly about what purpose can Facebook search play in making Facebook as a platform more successful.
But can’t you do both, since you do include results from Bing, along with results from Facebook?
In general, the last thing you want in search is for the user to hit a dead end right when they’re looking for something. You want to make sure that you give them something they can consume, something that informs the query that they just put in and I think that’s the big role that the Bing web results play in that.
A lot of times they, by themselves, are the best results, and a lot of times they sort of compensate for what is missing from our index, from what our own users have created, so I think all those things fit together really nicely. You can follow certain topics and say, I care about the US Open going on right now, what are my friends saying, a few friends are actually watching the game live, which is pretty cool to see happening. And then a lot of this stuff is sort of what is out there in the general web, and that’s what Bing sort of brings in to Facebook search.
How many people actually click on those Bing results?
In general, I would say they get used but the primary use case remains finding people, pages or following what your friends are saying about a particular thing.
Again, I think Bing is not front and center. (It) sort of adds value to what search is already doing. It definitely isn’t core. It definitely isn’t the most popular thing that people are clicking on but it’s definitely adding enough value that we actually put it in the product.
How do you think what you offer compares to Twitter search?
That, by itself, is a huge discussion — about what Twitter does really well versus what Facebook does really well. I think Twitter search is really interesting. I think they’ll have a lot of technical challenges to be able to — a lot of relevance issues to deal with. Facebook is the only one where at the scale that we are doing it, we sort of take all of the things you care about and are connected to and show you content from them.
I think Twitter still remains the pulse and the trends of what everybody is talking about, but Facebook has really always been about what we know about you as a person — what you care about and the things you’re connected to. I do see those things as being two very separate things. One is very personal, very much about, hey, did a friend of mine go and check out this restaurant and what did the person think about. That’s more the type of searches we are looking for than anything else.
You can take the “pulse” too.
I think if that happens that will happen organically and that’s a really good side effect.
What’s the ad situation?
(We have) the regular ads we show across the whole site. They’re not targeted toward the key word the user types in. We’re thinking about that — figuring out what to do in that space.
Search in general has been proven to be quite a monetizable product and definitely stuff we’re looking at being able to do more with. Today you will see that we show Bing ads … only in the U.S., which are based on the terms that the user types in so (they are) slightly more targeted than the regular ads we show across the site. So, yeah, definitely look forward to seeing more in terms of what we can do with that space.
For a long time Facebook search was pretty limited in its feature set. Why only now has that changed?
It’s been limited in the sense that, yeah, there’s a huge amount of potential out there. We are a pretty small engineering team.
What we realized is it is really interesting and useful for you to be able to search through your friends’ content. It is also an incredibly technically hard problem, which I’m sure you would understand: You’re dealing with 250 million users. So being able to actually independently go and say here are the only 100 users I care about, only show me content from them is actually a pretty hard problem.
So it’s something we’ve been thinking about for a long time. I think we sort of reached the stage where, you’re like, okay, let’s invest some time in this, (it would be) really awesome, really cool to get this out.
What you’re likely to see in the future is not so much brand-new, flashy stuff, but sort of just improving the stuff we already have out there: Make our ranking better, the relevance better. Today’s stuff is purely chronologically sorted.
Can you elaborate on what else could change?
It’s still sort of the initial phases so nothing I say is things we’re necessarily working on or have engineering resources to (do) it, but if you think about it, time is only one factor in what is most important. Who is saying it becomes really important too.
So, in general, our relevance is all about who are you as a person, what are the things you care about, what are the things you are connected to. So if you think about it we are already sort of by breaking out what your friends are saying from what everyone is saying so we’ve already carried out some sort of ranking, some sort of relevance match because we give that more prominence. So the idea is again how can we use your identity and what you care about to just rank the results and make the results more relevant.
Would you ever partner with a major search engine to feed them this data?
Anything that happens along those lines will be what we do with platform and what we do with APIs. Connect already allows a lot of that. People can simply add Connect to their sites and be able to pull in the stuff via the API. Obviously all the privacy stuff still holds true. So any sort of progress we make on that front will just be as the platform grows and as we innovate on the platform.
How will the acquisition of FriendFeed — which was known for its own search features — impact Facebook search?
I think they’re still coming in. They’re beginning to learn about the technology across the site.
Of course, they have some really good insight into search. Today we don’t use any of that technology, but I do see us looking more into that, getting to talk and work with the people who worked on search at FriendFeed. FriendFeed did work at a smaller scale than we did, but they just had really good search features. So I think there’s a lot we can learn from them.