Further to Shane’s great post on Facebook Graph Search Simplified I wanted to delve into what it really means for users and the impact that is likely to result across the web as we know it today. The Graph Search is still very much in beta and for now focuses on people, photos, interests and places. It should enable users to rely more on recommendations from their social group and take advantage of popularity of photos, posts etc to cut through the vast amount of posts and photos so that it’s easier to discover good content. For advertisers it will enable greater targeting capabilities based to try to reach users further down the purchases/engagement funnel.
Facebook has had it in it’s sights for years, collecting vast amounts of data on everything from users checking in at certain locations, to hot topics discussed (in a similar way to the Twitter trends). By working with Microsoft and the vast investment ploughed into it’s own search engine Bing, Facebook is now in a position to put the connections together in a way that it believes will be beneficial for users, by backing up the results with Bing web searches if Facebook has nothing to display itself. Google attempts to try and predict what a user is truly searching for by collecting user search data and providing ‘personalised results’. However, this is merely a drop in the ocean compared to the data that Facebook holds, meaning that the results may be more accurate and useful from the social media platform. If we take the artist ‘Jay Z’ as an example, a user may be searching for tour dates but Google could only draw inferences based on previous search history data held, which may result in music retailers taking up the 1st page of the results page. Facebook, on the other hand, could theoretically see that a user has checked in to concert venues before, has discussed gigs and likes the ‘O2 Arena’ page – this could lead to results being focused on any upcoming Jay Z concerts being held at the O2, which would be far more personalised (and useful for the user).
We’re not there today and for now it is claimed that the Graph Search will enable users to ask questions about which friends live in a certain city and combine it with interest-based data such as particular music tastes. Results would then be ranked by an intelligent engine that analyses social interaction, so that ‘closer’ friends would appear nearer to the top of the results. Great for users (just make sure you double-check your privacy settings as always), but even better for advertisers. For example, local businesses will be ranked in results based on ‘Likes’ from a user’s friends, in turn providing a social recommendation (the modern day word of mouth?) that the business can be trusted. This can also apply to national businesses, as we’ve seen recently with promoted posts, again using friends who have liked the page or company to help sell the product.
So where do we see this going? The huge data sets provide opportunities that have never been available before and there’s a few core areas that may be developed further over the coming years:
- Product Reviews: Facebook could look to take advantage of the existing ‘comments’ plugin on 3rd party sites (or look to rename to ‘reviews’) so that users could see if a friend had reviewed a purchase or product. For example this could be shown in results pages on the Facebook Graph search, or on another website where the review was written such as a retailer or dedicated review site. The benefits again here are word of mouth recommendations, enabling reviews to become more ‘personalised’ for a user rather than faceless, as certain opinions may be trusted more than others.
- Rich Snippets: In a similar way to Google including start ratings and author information next to results, Bing could shake the market up ,via it’s partnership with Facebook, by providing more relevant snippets on it’s own result pages, such as friends who liked a brand, reviewed a product or had visited a place.
- Location Targeting: Facebook and Bing could look to utilise data from mobile phone locations/check-ins and adjust search results accordingly. For example a user search for ‘restaurants’ from a mobile may be directed to a nearby establishment that a friend recently visited and reviewed positively. This would have huge benefits for advertisers if the data/snippets could be included with the paid listings, rather than just organic results (specifically for competitive terms in which a user’s friends have visited multiple restaurants from the example above).
- Traffic Generation: At iLikeOffers.co.uk we’ve developed our own personalisation system for offers, as well as a unique ranking system for discounts based on social shares. This means we encourage people to ‘like’, ‘tweet’ and ‘pin’ to their hearts content – currently providing a further traffic channel. If a user ‘likes’ an offer or brand then it will display on their timeline, as well as their friends’ news feeds. With the Graph Search, a user who has searched for ‘Tesco’ may see one of their friends has liked a Tesco voucher code offer on our site, which could result in an additional visitor if they decide to click on the link to get more details on the offer. For the web as a whole, it has huge repercussions in current traffic generation methods. Why? Well compared to the user seeing the ‘like’ in the news feed (which is pushed content – they may have no interest in Tesco whatsoever), if they are actively searching for Tesco it then becomes pulled content – they want the information. They are further down the engagement funnel and are a more ‘valuable’ visitor, perhaps shifting the focus from other traffic generation methods such as PPC.
This is the first time since Google launched that there is the possibility of a fundamental shift in the way web users find and access content – it’s worth watching this space like a hawk to see just what Facebook may have planned.Author's Google+