Steve Phillips, Cofounder and CEO, overseeing Zappi's global business, product and growth strategies.gettyMarket research was built for a world where brands made a small number of very big bets. For decades, brand building was anchored in high-reach, high-attention media. Marketers developed a big idea, validated a handful of major executions and invested heavily behind them. Today, marketing works differently. Media has fragmented, content has multiplied and brand growth increasingly depends on hundreds of smaller decisions made across channels, formats, audiences and moments.The world’s largest brands have shifted in this direction. Estée Lauder, for example, has spoken publicly about putting the vast majority of its media investment into digital, social and influencer channels. Less than a decade ago, many brands—including Estée Lauder—were doing something much closer to the inverse: investing heavily in a few statement media buys designed to build the brand at scale.This is the new “lots of little” reality. Brands are no longer managing one or two high-reach media moments. They are managing many lower-reach, lower-attention moments that collectively shape consumer understanding, salience and preference.That changes the job for marketers. It is no longer enough to validate the hero asset. Teams need to know whether the many smaller assets ladder back to the same brand idea, whether they work in context, whether they fit the audience and whether they are building something coherent over time.Insights Has Not Kept PaceIt’s not to say market research is totally archaic. In fact, it’s made enormous progress over the past 20 years. Studies that once took weeks with in-person focus groups can now be completed in hours. Data is faster, cheaper and more accessible than ever before. It’s progress that has helped insights make up ground as digital and programmatic media changed how marketers worked.But there is still a gap, and it’s widening again.When most media was mainstream, the biggest bets were the ones you validated. A television campaign, a major print campaign or a national outdoor buy carried enough investment and risk to justify deep consumer testing. As digital came into the fray, programmatic platforms gave marketers new forms of performance data: clicks, views, engagement, conversions and other signals that showed what people interacted with.This data was great at measuring engagement. It could tell marketers what people clicked, watched and purchased. But it was far less effective at measuring brand impact, brand coherence or whether the many assets within a campaign were reinforcing the same idea. That left a white space between what consumers did and why it resonated, a space that teams were left to fill with assumptions. Insights teams filled in that gap when they could. But practically, they could not validate every creative concept, every asset, every market adaptation and every personalized message. The system was built for the big decisions, not the many small ones that now ladder up to brand.The issue isn’t willingness. Most marketers would love to validate more. If they could test a digital ad, evaluate the personalized messages around it and understand how it would land with different audience profiles before spending media dollars, of course they would. The barrier is access.Consumer understanding has never been more valuable, but insights are still concentrated around too few decisions. For insights, the choice is simple: adapt to accommodate low-risk, high-volume plays or become further marginalized from the decisions that drive growth. Agentic Ecosystems Bring The Consumer Into More DecisionsThis is where AI can change the operating model.The most important use of AI in marketing may not be simply creating more content; marketers already have more content than they can manage. The larger opportunity is using AI to optimize, localize and validate content within existing workflows. In the next phase, insights will be enabled by an agentic ecosystem—a network of AI-enabled tools and agents embedded into research, brand and marketing. One insight agent might help develop a creative brief. Another might localize an asset for a specific market based on client insight data. Another might evaluate whether a digital execution is likely to reach, resonate and create a response. Another might monitor performance and recommend changes. The key is that these agents are trained and guided by connected data: historical research, brand guidelines, consumer response, campaign performance and competitive intel. If AI is the engine, these insights—which insights teams have in abundance—are the fuel. This data can power synthetic respondents which can immediately validate ideas at any stage of development—providing a temperature check in stride on assets that would otherwise go untested. That does not mean replacing human respondents. It means matching the method to the decision. High-stakes brand moments still require deeper human validation. But lower-risk, high-volume decisions need a different model. Synthetic approaches can help teams evaluate many assets quickly, identify which deserve further investment and learn what is likely to work before media spend is committed.This is especially powerful in a world where creative connects across products, brands and geographies. Brands need to understand not only whether individual executions work, but whether they work together. Do they share a coherent look and feel? Do they reinforce the same brand memory structures? Do they adapt to context without fragmenting the brand?For decades, the mission of market research was to bring the consumer into the boardroom. In a fragmented, AI-enabled media environment, the opportunity is bigger. We can bring the consumer into every room where decisions are made.The brands that win will be the ones that learn fastest, adapt intelligently and build systems that keep consumer understanding at the center of every decision.Forbes Technology Council is an invitation-only community for world-class CIOs, CTOs and technology executives. Do I qualify?