After last year’s Cambridge Analytica scandal, Facebook announced a major change to its ad targeting system. As of October, the company has permanently eliminated its “Partner Categories” feature, which had enabled advertisers

Whether or not Google winds up paying the fine, the writing is on the wall: Fines have so far proven insufficient to deter the world’s largest data collectors from gathering data in invasive ways. The market, however, could help.

A smaller, more accurate data market

Consumers could, of course, volunteer the data that companies are otherwise buying from third parties. But why would consumers or businesses want to shift to so-called declared data, zero party data, or information that users voluntarily provide through surveys, forms, polls, and apps?

For one, volunteered data is more accurate than the so-called inferred data that companies generate based on your online and offline transactions and behaviors. In a 2017 study of 107 U.S. professionals titled “Predictably Inaccurate,” Deloitte found that, across all categories, respondents believed 71% of their purchased data to be 0% to 50% correct. Even in binary categories like “owner of life insurance policy,” buyers found purchased data to be less accurate than a coin toss would be.

Accurate data matters because without it, personalization–something 90% of consumers find appealing, according to data broker Epsilon–is impossible. Fortunately, analysis tools have gotten good enough that customers can volunteer relatively little information compared to the multi-point profiles companies typically buy.

Eager for more personalization and consumer trust, companies are increasingly turning to zero-party data, according to Forrester. In a November report, the research firm said that 57% of consumers would share data in exchange for personalized offers. By next year, 15% of global brands will be collecting zero-party data.

[Image: courtesy of WayIn]
A number of marketing tech companies, including WayIn and PureProfile, are now focusing on user-supplied data. Jebbit, which calls it “declared data,” said that it recently worked with an online marketplace trying to grow the lifetime value of its auto parts customers. By sending out mobile surveys, it learned that just three data points could predict whether someone would make a purchase. It could then send coupons to borderline customers, increasing their likelihood of buying.