Facebook has developed a new advertising service designed to predict the future behavior of consumers, The Intercept reported Friday, after viewing a confidential document describing the offering.
The service uses FBLearner Flow, an artificial intelligence prediction engine the company introduced in 2016.
Facebook’s pitch to advertisers, according to The Intercept’s report, is that the technology enables companies to target people based on decisions they haven’t yet made, with the goal of influencing them to change their minds.
For example, if Facebook can tell a company which customers have started to think about jumping ship, that company then can put together a package of perks to reinforce customer loyalty and keep those customers in the fold.
The data the new service taps is aggregated and anonymized to protect user privacy, according to the report.
More Signal, Less Noise
Consumers could benefit from the new service by receiving advertising that’s more relevant to them, suggested Beerud Sheth, CEO of Gupshup.
“They aren’t receiving a wide array of messages that might not concern them,” he told the E-Commerce Times. “Instead, the ads consumers are getting are tailored to their interests and benefits.”
Predictive advertising can have its drawbacks, though, observed Charles King, principal analyst at Pund-IT.
“The longstanding pitch for predictive advertising technologies is that they help consumers cut through the noise of virtually limitless choice and focus on items they may actually want or need,” he told the E-Commerce Times.
“There’s some truth in that,” King said, “though the process also tends to homogenize consumer choice by assuming that what you want today — and will want tomorrow — is similar to what you preferred yesterday.”
Advertisers could benefit from Facebook’s AI-fired service in several ways.
“There’s no question that AI — when implemented correctly — has the potential to revolutionize advertising,” said Tod Loofbourrow, CEO of ViralGains.
“If brands know more about the tendencies of their target consumers, they can better tailor their ad campaigns to the people most likely to take action, optimizing their spend and improving their engagement rates,” he told the E-Commerce Times.
“They can even use this information to inform the development of their products and services in order to better fit the market,” Loofbourrow said.
The potential advantages to advertisers are “all to the good, but being associated with Facebook’s program may not be as attractive to advertisers today as it might have been a few weeks or month ago,” noted Pund-IT’s King. “Bottom line, this offering seems badly timed, at best.”
Facebook’s Fine Balance
Advertisers and data providers need to be careful about how they collect information about their customers, Loofbourrow warned.
“If consumers are willingly sharing their preferences with brands as part of a two-way dialogue, that’s one thing,” he told the E-Commerce Times.
However, “if data they thought was private is instead being shared — anonymously or not — with advertisers as part of ‘terms and conditions’ that are buried in fine print, ” Loofburrow said, “it continues the breach of trust that has already made itself apparent in the wake of the Cambridge Analytica scandal.”
Facebook needs to be careful about how it handles data following the improper sharing of information on 87 million users in the Cambridge Analytica episode, said Josh Crandall, CEO of NetPop Research.
“These are delicate times for Facebook. Use of any technologies to target Facebook users should be considered very carefully, especially right now,” he told the E-Commerce Times.
“Questions around where and how data are being used are front and center for influential, savvy users,” Crandall said.
“Rolling out AI at this point may be a short-term solution for quarterly earnings,” he continued, “but may be seen by some as another step towards long-term user distrust for the way Facebook respects user data and privacy.”