How A.I. can make digital advertising less creepy
We all know the shortcomings of digital advertising. You buy something online, and for days afterward you might be deluged with advertisements for the very same product on Facebook, Instagram, or wherever you digitally roam.
There are two big problems with our current digital advertising system. First, many consumers hate it because they feel like their privacy is being invaded. Who knew I wanted a garlic press, and how did they know it?
And, second, it’s not as effective as it could be. I’m unlikely to buy more of what I’ve just bought. So why does no one ever try to sell me a bag of garlic instead of another press?
Our entire digital ecosystem is about to change, and I believe for the better. Consumers will feel less spied on, and companies will be able to do a better job of anticipating needs.
In early 2021, Apple will require that users give permission to share their information when they use mobile apps. Marketers can continue to advertise, but they will no longer be able to tailor and assess their work in the same way, if users choose to opt out.
Cookies are little bits of data that are loaded onto your browser that let publishers, marketers, and others know what you do online. Without them, advertisers will have fewer ways to reach their potential customers.
In this new world, third-party cookies will be disallowed, but first-party cookies will remain. So if you are reading a newspaper site, the publisher will be allowed to use first-party cookies to see what you are reading and give you more of what you like. But third-party cookies, placed on that publisher’s site by, say, a tire company, will be forbidden.
The end of third-party cookies is a good thing, because it will make way for smarter technology. I believe that artificial intelligence is poised to transform the digital advertising industry—proving better value for consumers, advertisers, and publishers alike.
My company is partnering with a number of others to create a digital advertising ecosystem, based on A.I., that is open, transparent, and privacy-first—instead of walled and mysterious. IBM has a substantial operation built on helping bring A.I. to business, with IBM Watson clients spanning a variety of industries, including financial services, health care, and advertising. And we could benefit from the widespread deployment of A.I. in digital advertising.
A.I. can understand massive and diverse data inputs quickly, recognize patterns to make predictions, and continuously learn. A.I. technology, which includes machine learning, is already proving its worth in a broad array of contexts—like accurate weather forecasting, fast response times by government agencies to citizen inquiries, and cybersecurity systems that can monitor a welter of network noise to detect and disable threats. And so, A.I. should be able to help deliver what marketers have been seeking for years: a more precise way to aim their messages at the right customer at the right time.
Consumers want relevant, insightful advertising without the creepiness created by the cookie, which is essentially a tracking device. A.I. can recognize patterns and glean insights in ways that cookies cannot. It can see that a computer user lives in Washington, D.C., note that the weather is forecast to be humid and stormy, and pitch an ad for an allergy medicine—all without needing any personally identifiable data.
A.I. can pull from a range of data sets—such as zip codes, weather, retail membership cards, and shopping habits—to anticipate human needs. It can assemble and sort voluminous data sets, filter trillions of data points every second, and produce inferences about what a user’s next most likely transaction will be.
Some of the innovation in digital advertising in recent years has been superfluous, creating more opportunities for venture capitalists than true value for consumers, brands, and publishers. In many cases, brands are not getting much ad inventory for their money because an increasingly large portion of the spending is going to middlemen.
Sometimes, going back is a smart way to move forward. Advertisers, for example, may be interested anew in an old technique: the web crawler. It can see what is on your screen without needing to know who you are. If I’m on a tennis site, I might be served an ad for a tennis racket. The crawler does not get stored on your browser; it is just looking at the page you are on. Infused with A.I., the venerable web crawler can be a powerful inference engine without invading a consumer’s personal privacy. It’s but one example of how the future of digital advertising can work better for brands and consumers.
There’s a potentially big downside, though, to the end of third-party cookies: It will empower what we in the business call walled gardens—and that’s a bad thing for brands and an even worse thing for consumers. With cookies allowed only in first-party situations, platforms like Google, Amazon, and Facebook will have a walled garden advantage. They alone will have access to the data that advertisers want and need.
Increasing that walled garden power would be a disservice to users and to the digital economy, especially to small and upstart businesses. Worse, publishers risk making less money from advertising. And when that happens, they must live off subscriptions or shut down. The last thing we need is to lose independent publishers.
A.I. offers a way around the power of the walled garden and provides a positive new path forward for the digital economy. It’s time for consumers to benefit from a technology that is far smarter than the cookie.
Bob Lord is senior vice president of cognitive applications, blockchain, and ecosystems at IBM.