A few Questions About the New Facebook Search Graph
On Tuesday Facebook announced it was going into search. A long anticipated development with differing expectations. Are they going after Google? Or is it a completely different ball game? There are a lot of posts written about what it is, how it will work and what chances are for Facebook search. Heck, the first posts about SEO for Facebook search are already popping up.
So there’s a huge amount of data which will be accessible with Facebook’s new search. That offers major possibilities. But the amount of data accessible through a search engine is just one requirement for a successful search engine. After I learned about this new development from Facebook a few questions popped up. As everyone is talking about what it is I would like to talk about what I don’t know yet, but I’m very curious about.
Will they understand intent?
One of the major flaws of the old Facebook search was that you had to type exactly what you searched for, otherwise you wouldn’t get the results you were looking for. They didn’t control for spelling errors, they didn’t offer results for synonyms, they didn’t understand your intent. They just used exact matches for the query you typed. The problem with search however is that people do not always know what they are searching for or how to describe it exactly. You need advanced insights in languages and in searching behaviour to understand search queries. Will Facebook be able to get this right?
The power of Google is that they do understand what you want. They understand that when I search for pizza I don’t really want to know the history of the pizza. They know I want to know where the local pizza places are. How good Facebook can be at getting user intent defines how good search will work. My hopes are however pretty good. In 2010 Lars Rasmussen, a software engineer involved in the development Google Maps and Wave, moved from Google to Facebook. Especially Google Wave had some amazing features where language-based problems were overcome. With this intelligence Facebook search could be a lot better in understanding intent than we are used to.
Will the search results be relevant and good?
As Facebook states themselves in the announcement “the depth of personalization” is much higher on Facebook search. But is that something users want in search? Do we want a large set of general information tailored to our needs or do we want heavily personalized results? When I search for a local pizza place, do I want to see just the pizza places my friends like? Or do I want to see all the pizza places close to me enriched with what my friends like? My question is how deep will the personalization really be in Facebook search and can I choose how personalized it is?
The second question here is: are the results good? Do I find what I’m looking for? Facebook starts with showing people, photos places, business pages and interests. But when I search for a pizza place do I want to find the Facebook page of this pizza place? Not really. The Facebook page is heavily focussed on updates, not on information. So it comes down to what they show in the search results. Can they show me the relevant information there? Information like the address, opening hours, number of likes (from my friends) and, well, just one thing more… The website where I can order a pizza. Are they willing to send me away from Facebook? Possibly, they will also be showing web results amongst the search results, provided by Bing. But they will also prefer to keep me on Facebook. How far will they go in giving me what I want and how good can they give me what I need?
Do likes really represent likes?
One of the most important factors on Facebook to weigh search results will be likes. Where Google depends mostly on links Facebook will use likes to determine importance and to base personalization on. But are likes really a good way to weigh results? Do likes really represent things people like? When we look at business pages slightly more than 1% of fans of the biggest brands on Facebook are actually engaging with the brands (source). So do people really like what they like? Often people like things they don’t really like because they can win something. Likes will become even more valuable for businesses and pages, so prepare for more like & win promotions.
And a lot of things people actually like, they don’t like on Facebook. If you are satisfied with your dentist you are probably not going to like it on Facebook. Come on, who likes dentists on Facebook? What is the incentive? I know the most business pages I like are because I like the content they publish, not always because I like the business. How can Facebook make sure likes will actually represent what people like and how can they value those likes properly?
Do we need Facebook search?
A lot of questions about Facebook search. But the major question is do we really want to search on Facebook? Do we go to Facebook to discover new things? To find businesses? To find interesting people? Research shows we use Facebook for the following reasons:
- Staying in touch with current friends (91%)
- Staying in touch with family members (87%)
- Connecting with old friends you’ve lost touch with (86%)
- Connecting with others with shared hobbies or interests (49%)
- Making new friends (43%)
- Reading comments by celebrities, athletes or politicians (25%)
- Finding potential romantic or dating partners (26%)
It’s all about people, and only a little bit about discovering. It doesn’t really show us a real need for Facebook search, but it might create a whole new way of using Facebook we don’t think we need yet.
I’m really curious about the Facebook Graph Search. I believe it could actually work, but I also still have a lot of questions popping up when I think about it. What are you curious about? And do you think it will work for users?