A.I. Helps Online Ad Businesses Boom

For years, DribbleUp, a sports equipment company, spent its own time and resources figuring out whom it should advertise its basketballs and soccer balls to on Facebook. But for the last two years, it has placed those ads entirely through Facebook’s artificial intelligence tools.

Since then, DribbleUp’s sales have outpaced its marketing expenses. It has also started spending more money on Facebook.

The company’s experience speaks to how A.I. is reshaping the digital ad industry. Over the past three years, Google, Meta and other tech companies have used the same artificial intelligence behind chatbots to power advertising.

The emerging A.I. systems are helping companies automate their marketing. Small and large companies alike can now create ads, target audiences, bid for space and measure results. The process has made it easier for local businesses to develop campaigns as sophisticated as ones from corporate giants.

The technology is also making the ads more effective, tech companies and brands say. Meta and Google, the industry’s leaders, are using it to better match companies with potential customers, increasing the odds they are paid for their service.

More advertising money is flowing to tech’s biggest companies as a result. In 2022, the year ChatGPT launched, A.I.-related sales totaled $1 billion, but that rose to $35 billion last year, according to Madison and Wall, a consulting firm that tracks the industry. This year, sales are expected to balloon by 60 percent, to $56 billion.

“Google and Meta were already winning, and now, with these A.I. tools, they’re now lapping the field,” said Luke Stillman, a Madison and Wall managing director. The combination of the companies’ size, technological advantages and products means that in the age of A.I., “being bigger isn’t just proportionally better for Google and Meta,” he said. “It’s exponentially better.”

On Wednesday, Google and Meta are expected to report quarterly advertising sales growth that rivals the boom they experienced during the pandemic. The results are expected to build on last year, when Google posted a second consecutive year of double-digit advertising sales growth for the first time since 2019 and Meta’s ads business grew 22 percent.

That growth helps explain why the companies are spending hundreds of billions of dollars to build data centers for new A.I. systems. Google has committed to double capital expenditures to more than $175 billion this year, while Meta plans to spend more than $115 billion.

The soaring costs have contributed to Meta’s plans to lay off 8,000 employees, as it leans on A.I. to do work once done by engineers. Google has resisted making similar cuts and committed to hiring in key areas like A.I. and cloud computing.

A.I. has fueled a digital ad boom because the technology can instantly sift through large amounts of information. It is helping Google and Meta serve users more engaging content, which increases the number of ads the companies can display. It is collecting deeper insights into users’ interests, which improves the companies’ ability to target ads. And it is reducing advertising costs, which frees up money for bigger campaigns.

“Anybody that’s close to the space has seen a real shift change,” said Wesley ter Haar, chief A.I. officer at Monks, a marketing firm. “Technology is ready to meaningfully replace manual effort in our industry.”

There are downsides. Consumers may not only see more ads but also see ads that are more personalized, which can be unsettling. Advertisers have less control over how ads look, where they appear and how they perform.

“The process is more opaque, but more money is going into it because advertisers and users see benefits outweighing drawbacks,” Mr. Stillman said.

Where advertisers once bought ads using keywords or demographics in a fairly straightforward process, A.I. has begun automating those and other parts of the advertising chain. The result is a system that has become easier for companies to use, even as the technology behind it has become more complex.

The change begins with the way A.I. has increased the time people spend with Google and Meta. Because Gemini, Google’s A.I. system, is responding to many search queries with condensed and complete answers, people are doing more searches. Queries have doubled in two years, the company said.

For Meta, where videos have become a cornerstone of the business, the company has said A.I. has helped Facebook increase users’ watch time by double digits and broadened the reach of Instagram videos by dubbing content into different languages.

But the real business breakthroughs have come from targeting. It used to be that an advertiser would say, for example, “I want to target women in New York between the ages of 24 and 35.” Now it’s the opposite: Meta and Google are using A.I. to recommend customers the brands should be going after.

Last year, L’Oréal, the cosmetics company, used Google’s tools on 800 campaigns for 30 brands in 23 countries. It credited the tools with helping increase revenue for its NYX Professional Makeup.

“This only works if the models are good,” said Krassimir Karamfilov, Meta’s vice president of product.

Gemini is also helping Google parse people’s queries and evaluate advertisers’ marketing materials, so that the company can serve the most relevant ad in every search, said Dan Taylor, Google’s vice president of global ads. Irrelevant ads have declined 40 percent, which is a lucrative change for a business that is paid only when people click on an ad.

The A.I. tools are also taking over aspects of the creative and marketing process. Google has made it possible to change the messaging in ads in real time so that they better appeal to people searching online. The system, which Google calls “responsive search,” creates and tests ad copy before automatically serving a campaign that best matches a person’s search interest.

Loop, which makes concert earplugs, used Google’s A.I. tools last year to help it expand internationally. The system automatically translated ad copy, which the company’s marketing team had developed, for non-English language countries and placed ads abroad using keywords for specific markets, Mr. Taylor said.

A Loop spokeswoman said that the company used “real photography and real people,” but that A.I. had accelerated parts of its marketing process.

“Not only does it enable an advertiser to move quickly, but us to be able to move faster,” Mr. Taylor said.

Small companies have been some of the biggest beneficiaries of the new tools as they use A.I. to save money on designing the ads themselves.

Chris Wilhelmi, global head of data and media at Monks, the marketing firm, said customers were saving 30 percent on the cost of campaigns. On content, it can be as much as 65 percent. Many are reinvesting those savings into testing and experimenting with new ad strategies.

“By generating savings, you can fund new kinds of testing,” Mr. Wilhelmi said.

DribbleUp’s approach to marketing points to how things have changed. For years, brands evaluated advertising “like they did in the ‘Mad Men’ era,” said Ben Paster, DribbleUp’s chief marketing officer. Then came an obsession over analytics.

“Now everyone is approaching it as a technology problem,” Mr. Paster said.