I have written about some of the problems with Facebook Graph Search. First, I wrote about the problem of creating relationships between data where there aren't any. Secondly, I wrote about the problem of Facebook only searching within the meta data and not the content, and how people fill their profiles with junk info and thus distorting the result.
But there are also several interesting uses for Graph Search, one of which is how it relates to brands.
Let me give you a simple example: Here I have searched for "Clothing stores in San Francisco, California liked by women from San Francisco, California".
The great thing about this result is that it's very specifically targeted not only to a location, but also to a very specific group of people within that location. And in return you get a wonderfully ranked result of the best shops in that area, based on the women who actually live nearby.
This is just pure data porn!!
And just imagine the many other use cases similar to this. This is a great way for people to figure out where to go to buy something, but also a great way for people in marketing and sale to learn something about their competition.
Facebook Graph Search only looks at your page metadata. That is the title of your page, your location, and what category your page is in. So make sure you are not showing up as "corporation", but as a "clothing store".
Apart from that it's all about likes. The more people like you, and the higher an engagement rate you have (edgerank), the higher you will rank.
Facebook does not index any of your posts nor any of your photos. It only looks at the page itself.
Right now, Facebook Graph Search claims to be a 'natural language search engine'. It's not. Not even close. It is actually an assisted keyword engine. That may sound very technical, but it's an important distinction.
A natural language search engine understands what your question is and finds a result even if it doesn't match the specific words that you use. An assisted keyword engine only looks at the actual words and how they are ordered.
Why is this important you ask? Well, it means that unless you search specifically for categories Facebook has defined, you won't find anything. For instance, if instead of searching for "clothing stores" (a business word that nobody uses) you search for "fashion stores", you won't find anything.
Instead you are being directed over to Bing, who lists a department store museum as the top result.
Note: In comparison, Google's top result is "Best women's clothing stores San Francisco, CA" from Yelp.
You have to limit your search to the specific categories predefined by Facebook.
Another example: If you search for "Tablets my friends like who live in the United Kingdom", you won't find anything either. Instead, Facebook suggests that you find friends of a brand page called Tablet Celulares (whatever that is).
To find something, you have to search for the right category, for instance "electronics pages", which isn't very useful.
Note: Apple isn't listed because they are not doing anything with their Facebook page (zero rank score).
Another issue is that Facebook Graph Search is not searching within the content. It's only looking at the metadata (the graph data as they call it). What that means is that you can't search for, e.g., "Products people like from Nike". Facebook doesn't know what that is.
It suggests that you search for products liked by people who work at Nike (meta data), which is not the same thing. But even if you agree to this, you don't get to see any products. Instead, Facebook will show you the pages that are liked by people who work at Nike, within the page category of "product/service".
Yes, there are several very interesting things about Facebook Graph Search, but it's not a real search engine. It doesn't understand what you are asking (it's not a natural language search engine either like Apple Siri, or Google (voice) Search).
Nor does it look at the content, so it has no idea how things relate to each other. All it does is that it matches predefined fields in their databases, which is extremely limiting.
This also means that we are a very long way from being able to use Facebook Graph Search to learn what people think about a brand. We can search for 'page likes', but not for what people say or do, as we can on Twitter, Pinterest, Tumblr or Google+.
You should make sure your page data is up-to-date (title, location, category), and get people to like and engage with it (EdgeRank). This will give you the best chance of showing up. But it's going to take a couple more years before this gets really interesting.
What if you are a newspaper. Is there any interesting uses for you? Well, you can always search for "Newspapers liked by people who live in Italy", which will give you a nice list of newspapers ranked by their likes+EdgeRank. This is a quick way to learn about your competition and you should look at the ones at the top to see how they are using Facebook.
You can search for photos taken by people. For instance: "Photos taken at London 2012 - Olympic Park in July 2012", which will show you pictures from the place of the 2012 Olympics.
But the problem is that Facebook Graph Search is not very timely. You can't search for "Photos of Zuccotti Park from September 17, 2011" (date and place of the occupy Wall Street demonstration), because it will instead look at the whole month of September and not just the 17th ...and that specific search result finds no photos.
Facebook also wrote that journalists could search for "books read by journalists" or "books read by managers who work at facebook". But this isn't really that interesting ... and again, it won't find books per see. It finds Facebook Pages within the book category (it's not searching for the content).
The problem is that Facebook Graph Search is too wide, too general, and potentially too misleading to be used by newspapers on a daily basis. There might be a few special use cases, but this is nothing like a search on Twitter or Google+ where we can learn how people react to a live event.
It's likely to change over time (years), so play around with it. But remember to think critically about what you find.
Head over to G+ to comment and discuss this article.
Many publishers are starting to incorporate AI into their newsrooms, but most make the mistake of using AI to lower the quality of their news.
Climate change coverage needs a different focus, otherwise we lose our audiences
For publishers, a getting people to subscribe is only the very beginning of their conversion process.
Churn is something you have to manage long before people even subscribe
Many publishers talk about doing memberships instead of subscriptions. But what does that actually mean?
When publishers want to engage with their audiences, are they really unengaging them instead?
Only about 15% of the public pays for news, so how we do convert the remaining 85%?
Innovation doesn't happen in separation. It's a tool that you have to use to reach specific outcomes.
Founder, media analyst, author, and publisher. Follow on Twitter
"Thomas Baekdal is one of Scandinavia's most sought-after experts in the digitization of media companies. He has made himself known for his analysis of how digitization has changed the way we consume media."
Swedish business magazine, Resumé