Facebook Finally Lets Its Firehose Be Tapped For Marketing Insights Thanks To DataSift

Next Story

Lyft Hires Former Twitter Director Peter Morelli As VP Of Engineering

Twitter’s firehose of tweets has long been offered as a goldmine for businesses trying to understand how to improve or market their products, and now Facebook will allow privacy-safe peeks at its treasure trove, too. Today Facebook launched a new insights product called “Topic Data” in the U.S. and U.K. with the help of brand analytics leader DataSift.

Facebook explains that “Topic data shows marketers what audiences are saying on Facebook about events, brands, subjects and activities.”

For example, “A business selling a hair de-frizzing product can see demographics on the people talking about humidity’s effects on their hair.”

On days when everyone’s posting status updates about how frizzy their hair is, a brand could step up its ad spend knowing it’s the perfect time to reach potential customers. Sentiment, location, volume of mentions and words often mentioned alongside a brand can be pulled, too.

Out Of Datasift

Because much of Facebook’s data is private, unlike Twitter, offering Topic Data in a privacy-safe way is a top concern and might explain why Facebook waited so long to offer this functionality that brands have been begging for. To ensure personal info isn’t divulged, Topic Data is aggregated and anonymized, so brands can’t know or piece together exactly who said what. I asked Facebook and it confirmed that certain queries that might pull up personally identifiable data like home addresses will be banned. At least 100 different users have to match a query for it to be allowed.

Still, the idea that their private status messages to friends will fuel better ad targeting may irk some Facebook users. There’s no opt-out, and the only way to keep data totally private is to either set posts to be visible to “only me” or not post at all.

Screen Shot 2015-03-10 at 10.23.05 AM

To be clear, this isn’t a brand monitoring tool. It’s not designed to let companies see every mention of their business and try to respond. That wouldn’t work since data is anonymized any way.

Instead, Facebook’s product manager Matt Idema tells me marketers issue specific queries about “what’s being shared, what people’s opinions are about their brand, what are the trends, and anything that is top of mind.”

Brands issue the forward-looking queries through a third-party analytics provider that submits them to DataSift, which can run them against Facebook’s data. DataSift hands the analytics tool back anonymized statistical data about posts that match the query since it was issued that can be formed into charts and insights, or bundled with social analytics from other networks.

More examples Facebook gives for how to use Topic Data include:

  • “A fashion retailer could see the clothing items its target audience is talking about to decide which products to stock.”
  • “A brand can see how people are talking about their brand or industry to measure brand sentiment.”

privacy-first-lockMarketers can’t instantly turn a query’s results into ad targeting, but can set their ads to appear to people in similar demographics. Beyond just buying ads, Facebook hopes brands will discover things people want that they don’t make yet, and move to develop these new products.

When Twitter opened its firehose to this kind of analysis a few years back, it spawned an entire ecosystem of data interpreters, including Adobe Social, Brandwatch, Crimson Hexagon, Socialmetrix and DataSift itself, which was one of only a few companies allowed to sell the full Twitter firehose at one point. Facebook will authorize a limited, undisclosed list of tools it’s already working with to access its firehose through DataSift.

For now it’s only available to companies in the U.S. and U.K.; international expansion is expected eventually. “Privacy-sensitive Facebook has always trailed Twitter when it comes to monetizing data for business audiences so today’s Topic Data news is a big step forward and a fascinating evolution of their thinking” said analytics company Brandwatch’s CMO Will McInnes. “For the world of social intelligence, this is huge news.”

Surprisingly enough, Facebook is giving DataSift the keys to the castle for free in exchange for helping it rapidly break into the market with the startup’s tech and relationships. DataSift will charge analytics firms a fee for queries, who will then mark up the price and sell it to brands.

If brands find the data valuable, it could draw them and their ad budgets closer to Facebook. Until now, Facebook has largely been a black box inside a walled garden. Marketers didn’t know what was said inside. DataSift and Topic Data let them peer inside.