Facebook has access to a ton of user data and a new partnership is putting more of it to use for a select group of marketers. The new Topic Data offering gives marketers insight into trending events, brands, subjects and activities, based on aggregated and anonymized Facebook data.
Facebook turned to industry leader DataSift, proprietors of the Privacy by Design privacy-safe data processing framework, to help develop and scale the massive amount of topic data at their disposal. As a result, the brands with first access to Topic Data are existing DataSift clients.
The company isn't new to the social arena; both Twitter and Tumblr have used DataSift for filtered data feeds. The big difference, however, is that they were parsing publicly available data for those networks, while Facebook plans to tap into private posts with this initiative.
How Does Facebook Topic Data Work?
DataSift and Facebook created a new API called PYLON to share information back to brands while protecting the identities of the users from whom it was collected.
PYLON gathers data from both public and non-public posts on Facebook, then anonymizes the data. Likes, Comments, Shares, links and hashtags can all be collected for analysis.
The system then uses over 60 different attributes to sort the mass of data collected, including audience demographics like age and gender, as well as Open Graph data on topics and entities.
Facebook said in its announcement of the program, "While this type of data has been available from third parties before, the sample size was often too small to be significant and determining demographics was nearly impossible. With topic data, we've grouped data and stripped personal information from Facebook activity (not including Messenger) to offer insights on all the activity around a topic. That means marketers get a holistic and actionable view of their audience for the first time."
How Can Brands Use Facebook Topic Data?
The API will allow marketers to define which data they want to analyze using CSDL filters and run analysis queries on the data collected. You won't be able to see the actual posts, but will be able to see sentiment and an aggregate volume. For example, you won't be able to see that Mary from Philadelphia posted, "I just had the best Burger King burger!" But you will be able to see that 1,500 women ages 19-54 (for example) mentioned Burger King in a positive way.
But what would you do with this knowledge?
Plenty, says DataSift. Brand health, marketing optimization, revenue generation, customer experience, efficiency and innovation can all benefit, they say. Here are a few of the use cases they're highlighting on their website:
Share of Voice - Brands can use the API to monitor brand mentions (anonymized, of course). This is unique, as it gives brands an opportunity to see how often they and their competitors are being discussed in private Facebook posts.
Audience Research - The Topic Data API can help brands understand the types of people who discuss specific topics, by creating personas based on demographics information.
Campaign Effectiveness - Brands can see what audiences really think of their various campaigns.
Market Research - Discover what people are saying about your brand, products, competitors and the industry as a whole.
Topic Data can't be used directly for ad targeting, but can certainly inform your Facebook advertising strategy, driving messaging, targeting and offers based on its insights. Updated in real-time, Topic Data insights are available for a 30-day rolling window.
Tapping Into Facebook Topic Data
Currently, Facebook Topic Data is available only to those select brands in the US and the UK. DataSift promises they'll be expanding its availability over time, though they aren't specific about any next roll-outs.
For their part, Facebook said, "We are still discovering all the ways topic data can help marketers, so, to start, topic data will only be available to a limited number of DataSift's partners in the US and UK. (The data will be limited to those countries too, to start, and every partner DataSift works with for topic data is reviewed and approved by Facebook.) Our plan is to iterate on topic data thoughtfully and deliberately, and look to expand its availability at a later date."