As a trusted source for recommending everything from mobile phones to restaurants, word-of-mouth (WOM) remains almost peerless. Facebook’s new Graph Search tool aims to tap into the faith consumers have in WOM by delivering search results based on social connections and the interests and recommendations that those networks offer.
So, what is Facebook Graph Search?
Users will be able to search for terms such as, “restaurants in my hometown liked by my friends,” “museums in London visited by friends of my friends” or “articles on Marketing shared by Marketers” and many more combinations and topics. What’s more, each search result is bespoke, compared with keyword searches which return almost identical results for everyone.
For Marketers this opens the door to a room in the house which the omnipotent Google doesn’t yet occupy. Users want search to be more social and Graph Search has an abundance of consumer profile data which Marketers can use to help determine what their customer’s preferences are and tailor their Marketing activities and online content accordingly.
Mark Zuckerberg has described Graph Search as the ‘third pillar’ of Facebook, alongside News Feed and Timeline. Thanks to Facebook, we now associate the thumbs up symbol with its ubiquitous Like button, which is clicked by users the world over to applaud any manner of performance, product event or comment.
If, however, you have doubts about the reliability of Likes, the sturdiness of the pillar’s foundations may come into question. As even occasional Facebook users can observe, Liking something is often done arbitrarily and in some cases tinged with irony.
However, researchers from Cambridge University have revealed that Likes can indeed be an accurate indicator of personal attributes and preferences. The study, printed in the journal PNAS, is based on the online activity of 58,000 Facebook users who submitted their Like history, demographic profile and the results of psychometric testing. The Likes were then fed into an algorithm and verified with the volunteers personal data.
The findings showed an 88% accuracy rate for determining male sexuality, it was 95% accurate in distinguishing between African-American and Caucasian Americans and 85% accurate when differentiating between Republican and Democrat supporters. It was also 75% correct when predicting drug use. While some will be shocked at the extent to which information of a highly personal nature can be revealed, there were also some more light- hearted, if somewhat ambiguous results. Take for example a penchant for curly fries correlating with high intelligence, or liking the Dark Knight film indicating a low Friend count. Fans of the humble chip and Star Wars box sets be warned.
Better marketing intelligence
Having the ability to accurately infer customers tastes, preferences and traits may allow Marketers to improve their products or services and attune them to their customers. The authors of the study also suggest that psychological data could be used in marketing materials. For example, an online insurance advert could choose to emphasise security or potential threats, depending on the emotional stability of the user, a technique which would surely raise some eyebrows.
Graph Search is just being rolled out, so it is too early to herald it as a competitor of Google. What we do know though, is that the more Likes an organisation receives, the more chance it has of showing up in Graph Search. Of course, that’s no reason to stop link building,but Google might have to make some more floor space available in their house in the near future.
[minti_button link=”http://marketingmavens.io/free-marketing-how-to-guides/essential-guide-to-setting-up-social-media/” size=”full” target=”_self” lightbox=”false” color=”color-2″ icon=””]Download Our Essential Guide To Setting Up Social Media[/minti_button]