The trend toward greater focus on privacy issues has been going on for some time and is starting to come to a head. More restrictions on the sharing and merging of data on individuals has been leading to advertisers to look for effective ways to target and reach consumers, including using the use of behavioral targeting supplemented by the use of artificial intelligence (AI).
At a time when privacy regulations are sometimes fragmented and confusing but changing, it is critically important for marketers to monitor changes in the regulatory environment. Against this backdrop, I interviewed Sheri Bachstein, IBM’s Global Head of Watson Advertising to get her insights and predictions on the future of privacy regulation and how it will affect advertisers, particularly as regards the use of AI and came away with three major takeaways:
1) The need for standard federal regulation of data privacy in the U.S. is pressing.
The European Union’s General Data Protection Regulation and the California Consumer Privacy Act are already leading to the devaluation of traditional third-party cookies and the way many advertisers do business. Yet, the lack of uniform regulation creates a significant gray area for companies who want to scale and use AI in targeting, creating problems for technology companies and marketers.
Bachstein believes that federal legislation as opposed to a patchwork of state laws is needed in order to allow for standards that allow marketers to operate effectively while protecting consumer rights. Citing lack of agreement on what “sensitive data” means to different constituencies, she asserts: “We feel that data privacy principles such as transparency, user choice, as well as overall accountability should be taken into greater consideration when designing policy that is consumer driven. When you talk about “sensitive data” that is a great example of why we need industry standards. We all need to have the same definitions for types of data so that industry players work with clearly defined terms that are consistent across the U.S.”
Bachstein also notes that industry needs to unite behind a standard playbook and the effort to get federal regulation should be include an array of industry partners, councils, and big tech companies collaborating with lawmakers in order to incorporate multiple viewpoints and ensure that the legislation is effective across the entire ecosystem.
2) Artificial intelligence is part of the solution to balancing privacy with the consumer’s right to choose; Unified IDs alone are not likely to be enough.
Even before the pandemic, there was growing consensus that users should have the ability to control their sensitive data and who they share it with. Moreover, there is emphasis in regulatory discussions on consumers knowing where their data is going and how it is used. At the same time, most consumers prefer to have personalized experiences online, including when they receive advertising in the digital context. Old style “spray and pray” ads have little utility for either marketers or consumers. As such, an alternative targeting solution that does not use data the consumers believe to be sensitive is needed.
Because of its ability to narrowly target messages based on behavioral online data (in some cases combined with consumer approved data that is less sensitive) artificial intelligence has the potential to help get advertising messages that the consumer wants to see, or, at least, does not find invasive. It is also the case that the use of AI combined with informing consumers on how data is used can help lead to a situation where consumers understand that marketers are using data for positive purposes.
A key to the promise of AI in advertising, says Bachstein, is its ability to use data to predict future behaviors without relying on consumer-level data. “AI is a great path forward in this era of privacy because it can work with data to predict future behaviors, while simultaneously delivering valuable insights,” she states, “This allows publishers such as IBM’s The Weather Channel, to create relationships with their consumers without the need to know specifically who the consumer is.”
While some have suggested unified ID programs such as the The Trade Desk’s Unified ID 2.0, which is a pool of industry collaborators who seek to create to new, standardized ways of collecting user data without using third-party cookies, instead relying on audience data collected directly from the publisher, Bachstein does not see it as being able to replace what was available via third-party cookies. “While we welcome industry collaboration and view this solution as progress, it’s important to note that there will be challenges that are dependent on which path the big tech companies will pursue,” she says, “Unified IDs may offer a partial fix, but we could ultimately run into the same privacy constraints moving ahead. While Unified IDs are a step in the right direction, the solution alone isn’t enough to replace what marketers will lose when third-party cookies are phased out and mobile identifiers are changed to opt-in.”
3) The use of artificial intelligence is poised to grow, lead to additional innovation, and drive the future of digital advertising.
Bachstein and IBM Watson estimate that only 25% of global companies understand the true value of AI but that this will grow as marketers realize its value. In addition, consumers have not historically understood the benefits of AI and in part due to its futuristic image are nervous about how its application affects their everyday lives. IBM Watson believe the onus is on industry to explain the benefits in a delicate way to get across that AI will be a force for good.
As cookies have the drawback of only tracking the past, AI’s nature of being rooted in the future and leveraging data can deliver important insights without the publisher (e.g., IBM’s The Weather Channel) needing to know who specifically the consumer is. Bachstein summarizes the situation regarding adoption as follows, “As with any new technology, the pace of adoption and the consumer’s level of comfortability is ultimately tied directly to the education and experimentation of the technology itself. While there has not been a true resistance to AI… today it is not widely used across the digital advertising ecosystem. Adoption will take time for some, while others will see the benefits immediately. We need to ensure that AI is no longer perceived as a buzzword, but rather a tangible solution that can deliver real outcomes while keeping consumer privacy intact.”