AI may be capable of extraordinary things, but for marketers, its biggest opportunity right now is decidedly less exciting.

“You've heard a lot of great things about AI so far today, but I wanted to share a secret with you, which is that actually AI is pretty boring,” said our analyst Nate Elliott at EMARKETER’s Future of Digital summit. “And I want to be clear: Not only do I not think that boring is a bad thing, I think that boring may well be AI's superpower.”

Rather than waiting for AI to reinvent advertising, marketers may find more immediate value by applying it to the repetitive, data-heavy work they already do.

AI as a channel has room to grow

AI is changing marketing in two main ways: as a channel for reaching consumers, and as a tool marketers use to do their jobs.

The first has attracted much attention.

“More than a billion people every month use ChatGPT and Google Gemini and even Google AI mode,” said Elliott. “This has grown faster than any tool or platform or channel we've ever tracked at any marketer.”

But usage remains small compared with other media. The average US consumer spends just 16 minutes per day using AI, less than 3% of a 14-hour media day, Elliott said.

AI ad spending tells a similar story. US spending will reach $32 billion this year and more than double to $68 billion by 2030, according to EMARKETER forecasts cited in the presentation. But most of that spending is traditional paid search appearing alongside AI-powered results.

Advertising within AI chats and conversations will account for about $5 billion this year.

“Advertisers in this country are going to spend more on roadside billboards in 2026 than they're going to spend advertising inside of AI,” said Elliott. “AI as a channel has captured our industry's imagination. It's already grown remarkably quickly, and it's going to continue to grow in terms of users and dollars and importance over the next few years. But right now, these numbers are kind of small.”

The less flashy opportunity is already here

AI’s role as a marketing tool is considerably further along. Only about a quarter of surveyed marketers have tried AI as an advertising channel, while 74% said they use AI as a tool for their marketing jobs every day, according to Elliott.

Those applications span audience research, strategy and planning, creative, media buying, and measurement. But AI is still far from fully embedded in marketers’ workflows: At most, 40% of respondents said AI was regularly incorporated into any individual marketing activity, Elliott said.

Where has AI made the most progress? Audience research and campaign measurement, the less glamorous parts of the job.

“Good audience research and good campaign measurement are absolutely critical to successful marketing campaigns, and we've ignored this for far too long because it's not the interesting or the sexy part of what we do,” Elliott said.

“These tools were not built to replace human ingenuity and creativity; they were built to connect the dots,” he added.

Specific AI jobs beat moonshots

Marketers looking to demonstrate AI’s value should start with specific problems rather than sweeping transformation projects.

“The best thing you can do is focus on a handful of specific tasks, and set yourself specific numerical goals for those tasks,” said Elliott.

A Carnegie Mellon University study cited in the presentation examined the financial results and AI performance of 500 public companies. AI hiring, overall investment, employee adoption, proficiency, and AI maturity benchmarks did not correlate with revenue growth.

What did correlate with revenue growth was what researchers called “narrative concreteness,” how specifically companies identified AI use cases and the numerical targets they hoped to achieve.

Edelman is taking that approach by identifying 300 processes used across its client engagements and examining them individually for opportunities to use AI, Elliott said.

“It's not the most exciting work, but it's going to be incredibly important to their business and to their clients' success,” he said.

General tools produce general results

Finding specific AI use cases is only part of the equation. Marketers also need tools built for the job.

Among marketers using AI for media strategy, planning, and buying, general-purpose AI chatbots were the most commonly used tools. The same was true for marketers using AI for creative work.

“No wonder marketers say the number one reason they ignore AI recommendations is that those recommendations feel generic,” said Elliott. “Of course, they feel generic. We're using the same AI chatbots to plan our marketing campaigns that we use to shop for Christmas gifts.”

The more immediate AI playbook is much simpler. Marketers should identify repetitive or annoying tasks, choose technology suited to those jobs, and start putting it to work.

“They want this to be the tool of their professional lifetimes, and I think there's a good chance that in two or three years, it will be,” Elliott said.

For now, though, the opportunity is more practical.

“If you want to use AI successfully right now, the best thing you can do is identify the most repetitive and annoying tasks inside your organization,” said Elliott. “Make sure you have the right AI technologies to help you accomplish those tasks, and just get started.”

Watch the full session.

This was originally featured in the EMARKETER Daily newsletter. For more marketing insights, statistics, and trends, subscribe here.