$FB Graph Search - Myopic Myths: Busted.
I’m saying this for the third time. Facebook will do search. And [more importantly] they’ll do it well. Yes, they announced Graph Search in mid January, but since then the naysayers beat up the product with myopic criticism of the beta release. Some months have passed, but I figured now that everyone is focused on beating up Facebook Home instead, it’d be as good a time as any to answer back….
[[MORE]]Myth #1: Nobody uses Facebook search.
Facebook is still amonth the top three most visited properties in the world on any platform (desktop or mobile) by any measure with more time spent, engagement, etc. Sheer volume alone would suggest people had been clicking on the search box on the top left corner of the screen (if only by accident). I’ll concede I don’t have those numbers, but Facebook does and it’s unlikely they’d put out a product that wouldn’t leverage that behavior. Moreover, we had all been using Facebook’s search all along and likely more than we realized. Traffic to profiles comes from clicking on newsfeed and timeline links and from typing someone’s name in the (original) search bar. Yes, the original search product was terrible (you could really only search for people and even that wasn’t great), but that doesn’t mean people weren’t using it. Rather, people were using it despite how bad it was. And now it’s better. So I would venture that it will be used more.
Myth #2: Graph Search is based on only likes.
This is perhaps the most significant criticism of graph search that I’ve encountered - that search results are based on “likes” and likes are incentivized clicks (i.e., crap) and therefore the data is dirty and unusable. A well thought out discussion of Facebook’s dirty like data is here. I don’t deny that advertisers spent lots of money to get “fake” (or incentivized) likes that do not reflect a user’s true preferences. But….
Relevance algorithms should (theoretically) be able to account for “dirty likes.”
It’s not as if Google search isn’t gamed via SEO and other techniques.
It is *highly* unlikely that Facebook will use likes as the only or primary measure of relevance.
It’s the last point that strikes me. Likes are data Facebook can use (be they incentivized, crap, or otherwise), but they are hardly the only data Facebook can use. Facebook posts seem to be filled with lots of links as well - I can’t see why those links couldn’t be indexed similar to how Google indexes links on web pages to determine search relevance via PageRank. And why stop with likes and links? What about pages, mentions, music, products, tv shows, books, etc. All of which are effectively represented by links at some level anyway. While Facebook’s data is hardly perfectly structured, it’s better structured than we realize and getting more so.
Myth #3: Graph Search has no business model.
This is just crazy town. The missing, um, link for exploiting social networking as an effective advertising medium has been the lack of intent. “They” say that when people search for something via Google they are searching with (purchase) intent. “They” are right. But “they” say this has been lacking on social networking. Well, now there’s Graph Search and that means that people will begin to search with intent to purchase. Important reminder: people did not originally use Google to search with purchase intent. That evolved. As will Graph Search. Google taught us to search for information on the web - purchase intent followed. Facebook is teaching us to search for information on social media (well, Facebook at least) - first for people… then for photos… then for music… then for games… etc. This product will evolve too. In fact Graph Search doesn’t let people search their own posts for information “yet” (which would seem like an obvious place to start), but I surmise based on the image below that this is on the todo list.
Myth #4: The privacy issue.
People said the same thing about the Web, cookies, Google, and all sorts of social media. Privacy is dead. End of story.
Funny thing is….
Social search is already here. Whatever you may think of foursquare and/or whether it even survives, I believe what they have built with their Explore feature represents the closest thing to the future of social (and mobile) search that we have today. The results returned are based on not only static information (e.g., “what is in this particular spot?”) and not only on location and other data (e.g., “what is near me?” whatever “near” may mean), but also on where I have been and where my friends have been (i.e., the social component). That is, my social graph impacts the relevance of Explore (i.e., search) results. And there’s no denying there is purchase intent when people Explore. Again, many believe that foursquare will not make it (I am not one of those people), but foursquare’s current apparent struggles are the result of a lack of growth and scale - not a lack of business model or utility. Facebook has growth and scale. Oh and so does Twitter (so everything I just said applies to Twitter too).
To be fair, the current Graph Search product is indeed lacking in features, functionality, and utility (some would even argue it sucks). But it would be extremely myopic to not see its direction. And a myth to not appreciate its revenue potential.
[Note: I am “slightly” long $FB and picked up a little Twitter in a secondary market transaction.]