Facebook Graph Search is Facebook’s latest big change that has the tech world buzzing, but don’t get too caught up in the hype and FUD yet. It’s generally an improvement to Facebook, the kind that you would welcome in open arms. In specific, however, it’s another worrisome and possible invasion of privacy.
Experts are still analyzing the feature and mulling its consequences, but for now, let’s try to understand what it is based on what we know for now and what it means for the average online marketer.

• How Graph Search Works

The best way to explain Graph Search is think of it in terms of Boolean searches. Without getting too technical, to make use of Graph Search, you have to use certain words and phrases in a certain order.
You already know one of those phrases: “friends of my friends”. You type that in and Facebook will list them all for you, and you can refine it further; “friends of my friends who are single”, “friends of my friends who live in Minnesota”, “friends of my friends who are married and live in Ohio”, etc.
Facebook Graph Search gets really useful when you come to searches like “Music my friends like” or “Restaurants near me that my friends like”. These searches bring to mind the searches made in iTunes, Amazon, Netflix, etc. Those searches drive buys and bring ROI, although many of those websites have a discovery component as well.

• The Concerns with Graph Search

Facebook wants you to believe Graph Search is not going to invade your privacy. It only gives search results from data that users have made public. Conceivably, if you don’t want a secret sanctuary made public, or some other scenario like it pops up, you just need to make sure that info was shared privately and it won’t be covered in Graph Search.
However, the problem is much of the Facebook data on you is obscured. What this means is as you accumulate likes, shares, and other activity, individual actions get lost in that mass of activity you’ve done. We also need to account for activity acquired through interactions you did not know shared your information (liking page or site.) Facebook faced heavy criticism for a newsreader system and had to shuttle it, but nowadays you can use Facebook to do so many things on other sites; leaving comments on a blog post, logging into a site to use their service, or even joining another social network. Even seasoned tech experts can often take for granted where they’ve used their Facebook accounts, and not all that data is easy to find, if it can be accessed at all.
And even in cases where the data is easy for the user to find, not everyone is tech savvy enough to understand how or even why they should protect their data. This basic factum is the reason privacy advocates keep objecting to all these different changes implemented on Facebook. A user might be too young (and ironically, not supposed to be using Facebook), too old, naive or just plain misinformed.
Essentially, these functions open up Facebook user data to abuse. Concerns that tech websites have already identified include finding private information, data mining, and even worse, social engineering. Just for a quick rundown:

• A few sites, questionably or not, used the feature to find which Facebook users liked pages that would imply that they were sexist, racist, neoNazi, etc., although that may not jibe with real life. In many cases, likes may have been meant ironically, with pages themselves being facetious and parodying in nature. There are also cases where pages are misidentified based on keyword: a rape crisis center using all the wrong keywords could make it appear that its likers are for rape.
• Data mining is nothing particularly new; what it is is using the available data for market research. The concern here is that we as individuals are not informed when our personal details are used for said market research. Some would argue that everyone in Facebook had to agree to certain provisions when they checked the tick box saying they read the Terms and Conditions, but that may not really hold water in real life. Facebook has no guarantees all their users actually did go through those Terms and Conditions, and more importantly, those same Terms and Conditions can be rendered invalid or abusive in court if they are found to infringe on consumer rights. (Note: the author does not take legal responsibility for expressing this opinion. Please contact your lawyer if you need advice regarding the legalities of using Facebook.)
• Lastly, social engineering, which was highlighted last year when Wired staffer Mat Honan’s entire life was hacked , could be enabled by the Graph Search. Graph Search makes it easy to find and verify the very same personal details needed to usurp someone’s identity. With this ease of acquiring data, the possibility opens up to social engineering on a wider scale.

• The status of Graph Search now

With all the speculation, it’s still too early to tell if Graph Search will be as dangerous/innovative as expected. Since the feature is still in beta, it has not had the time to have mined all publicly available data yet. So, the few beta testers using it now may be surprised to find even more of their data exposed later.
Also, all the marketing potential of Graph Search is still just potential; we have yet to see Facebook’s eventual game plan in monetizing the new search features. One hopes they will balance out their user’s and advertiser’s needs.
 

• Facebook Search and Bing

It is also worth noting that Bing was not involved with Graph Search, but they also help power Facebook’s search. Bing has closely matched Google and fought hard to steal some of Google’s marketshare, although Bing’s algorithms remain noticeably less capable. Facebook is not holding their breath for Bing to catch up, and with Graph Search they are seeking to define what Search should be for them.

• Conclusion

As online marketers, it’s important to stay on top and be informed on the status and progress of Facebook Graph Search, but it’s too early to tell if it will actually be useful, or if overwhelming privacy concerns will force it to be shuttled or gimped.