AI search is quickly changing how consumers discover and evaluate brands, but marketers are still figuring out how to measure, and improve, their visibility. Most marketers are still experimenting, often building on existing SEO strategies.

But AI search requires marketers to think beyond where their websites rank. AI-generated answers can draw on brand websites, publishers, social platforms, reviews, affiliates, and community conversations, making visibility a much broader challenge.

“You can't just look at where site ranking is today,” said Richard Mastriani, JPMorgan Chase’s head of SEO/AEO, in a panel discussion at EMARKETER’s Future of Digital summit. “You have to look at where the site is being mentioned, or how the brand is being mentioned. And then, what's behind driving that mention?”

AI visibility requires a new measurement framework

Traditional search metrics aren’t disappearing, but they’re no longer sufficient.

AI platforms should currently be viewed more like brand channels than performance channels, according to experts at the event. That means marketers need to consider several measurements including inputs, brand visibility, and eventual performance.

“Your brand measures are going to be the share of voice, visibility rates, citation sentiment, citation accuracy,” said Michael King, founder and CEO at iPullRank. “And then your performance metrics are going to be what they've always been. How much referral traffic do you get, and then how does it perform?”

Marketers shouldn’t put too much weight on any single number, however. Instead, brands should look for persistent trends and focus on areas where visibility can ultimately influence business outcomes.

SEO is the starting point, not the entire strategy

Many of the fundamentals of traditional search still apply to AI discovery. Sites need to be crawlable, content needs to be retrievable, and brands need authoritative content that directly answers consumer questions.

“Quality content is going to win, and I don't think that's going to actually change,” said Mujahid Robinson, Profound’s manager, agency engagement, at the panel.

But AI search raises the bar. Brands aren’t just competing against other websites for visibility, they’re competing against AI systems that can synthesize information and answer questions themselves.

“Classic SEO is the price of entry,” said King. “And that just means that you are potentially eligible. It doesn't mean that you are going to automatically perform.”

Marketers need to think about the entire content ecosystem, he added, because success increasingly requires marketers to expand their focus beyond their owned websites.

That means AI visibility could require greater coordination among SEO, content, PR, social, creator, and other marketing teams.

Don’t chase volatility

AI search may also shorten marketers’ experimentation cycles.

“In the case of AI search, things happen a lot faster than what we saw in classic search,” King said. “I would give a 30-day window and then have it tied to whatever KPI matters for you in your business.”

But one positive result isn’t necessarily a reason to scale. Marketers should first determine whether an improvement is repeatable, persistent, relevant to their audience, and connected to a specific action they took.

“Visibility, I think, is too broad,” said Robinson. “If you are showing up, what is the actual thing that's being said? Is it something negative? Is it something positive?”

That distinction matters because brands don’t control many of the sources influencing AI answers. Reddit discussions, YouTube videos, publishers, creators, reviews, and other third-party content can all shape what AI platforms say about a company.

Brands may need to engage with those conversations rather than simply optimize their own websites.

“If everyone's complaining about a policy, like that's not gonna get changed by a tactic,” King said. “Negative AI sentiment can therefore be a marketing problem, or a signal that the underlying customer experience needs fixing.”

There’s no shortcut to AI visibility

For all the attention around GEO and AEO, the panelists cautioned marketers against tactics promising an easy way to manipulate AI results.

“I'm skeptical of anything that's a shortcut,” said Robinson. “I'm doubling down on anything that is tried and true that has worked over the long term.”

That includes building authoritative content around consumer intent and questions, maintaining strong technical foundations, and increasing a brand’s relevance across the broader web.

Technical performance also matters. Because AI systems can retrieve information in real time, slow websites risk being overlooked.

“Page speed is something that's really critical here because a lot of these systems again are pulling in real time,” said King. “So if your site is too slow, they'll just skip it and keep going.”

The takeaway for marketers: AI search may introduce new metrics and new sources of influence, but it hasn’t eliminated the fundamentals. Brands should make themselves easy for AI systems to find, understand, and verify across multiple sources, and resist chasing every new optimization tactic that emerges.

Watch the full session.