Consumer discovery is no longer just about where people go, but about how their thinking is shaped before they get there. As AI-driven platforms reshape decision-making, the challenge for leadership teams is understanding which parts of the consumer journey they can still see, and which parts are now being shaped beyond their line of sight.
Over the past year, early behavioral signals suggest something more fundamental is changing in how people discover, consider, and choose products and services online. These shifts point to a reconfiguration of where influence happens in the consumer journey—increasingly upstream of traditional touchpoints, before leadership teams have visibility.
Traditional digital discovery assumes that intent is largely formed before someone arrives at a search engine, a marketplace, or an app. By the time users engage with a brand, digital systems are primarily responding to decisions already in motion.
What we're beginning to see with large language models challenges that assumption.
Conversational interfaces don’t just surface options. They increasingly participate in the thinking itself. Consumers spend time inside these environments clarifying what they want, comparing trade-offs, and refining preferences before they ever encounter a brand directly. In some cases, the direction of a decision may already be taking shape upstream of any measurable interaction.
We’re starting to see this play out in how large platforms are experimenting with deeper integration across content, commerce, and recommendation. When companies like Netflix and Spotify explore closer alignment between discovery, content, and monetization, the significance isn’t the partnership itself. It’s the signal that discovery, consideration, and engagement are increasingly designed to happen within fewer environments, for longer, and with less friction between “thinking” and “doing.”
Similarly, when conversational interfaces begin surfacing products or services directly, they are not replacing existing channels overnight, but they are testing whether influence can be exerted earlier in the journey before brands have visibility. Already, our data shows that 1 in 25 Amazon sessions are preceded by a ChatGPT session—a figure that has grown 60% in just six months.
For organizations that rely heavily on first-party data, this raises an uncomfortable question: how much of the decision-making process are we no longer seeing?
The early signals are most visible in higher-involvement categories, where uncertainty is higher and consumers benefit from extended exploration. In these cases, traditional analytics may capture activity after influence has already occurred. That doesn’t make existing data wrong, but it does suggest it may be incomplete in ways that matter strategically.
At the same time as discovery is evolving, organizations have become increasingly sophisticated at measuring behavior within their own digital ecosystems. First-party data has never been richer. Teams can see, in detail, how users move through their apps, platforms, and owned environments.
The problem is not visibility, but where that visibility stops.
As more of the decision-making process happens elsewhere, leadership teams may feel well-informed while still missing important shifts in the broader competitive landscape. Changes in consumer behavior that occur upstream, across platforms, or outside owned environments can remain invisible until their impact is already reflected in performance metrics.
This is compounded by the pace of change. Quarterly reviews and traditional reporting cycles were built for environments where behavior evolved gradually. Today, shifts in discovery, consideration, and preference can take hold in weeks or even days. By the time they appear in reports, the competitive context may already have moved on.
The challenge for leaders is reconciling two truths at once: having more data than ever, and still lacking a complete view of how decisions are being shaped across the wider ecosystem.
Given this shift, the most valuable role of data is helping leaders ask better questions.
How are AI-driven discovery tools influencing consideration before consumers reach owned channels? Which parts of the journey are becoming less visible, and which new signals deserve attention? When large language models surface products directly, do those interactions meaningfully change behavior, or simply repackage existing demand?
Cross-platform data can answer these questions. For example, our data shows that shoppers arriving on Amazon from ChatGPT are more intentional: on Black Friday weekend, they spent longer shopping, viewed more products, and converted at higher rates than those coming from other channels.
Recent experimentation across commerce, content, and platform ecosystems suggests that many organizations are actively probing these questions in market. Early partnerships between large language models and commerce platforms are less about immediate scale and more about learning: can AI-led discovery meaningfully influence choice, not just redirect traffic?
The open question for leadership teams is not whether these initiatives will succeed in their current form, but what they reveal about where influence, attention, and decision-making are beginning to concentrate, and how quickly that influence might shift.
As AI continues to reshape discovery and decision-making, the greatest risk for organizations is not being wrong about where things are heading. It is assuming that existing models of consumer behavior still reflect how decisions are actually being formed.
When influence moves upstream, and thinking happens inside environments organizations don’t fully observe, the question is no longer whether change is coming. It is whether leadership teams will recognize it early enough to respond deliberately, rather than reactively.
The organizations best positioned to navigate this shift will not wait for perfect information. They will pay close attention to emerging behavioral signals, test assumptions quickly where risk is manageable, and remain willing to adjust course as patterns become clearer.
In a landscape defined by speed and complexity, the ability to see consumer behavior changing in real time may matter more than making confident assumptions about where markets will ultimately settle.
These patterns are still taking shape, but the questions they raise are worth discussing. If you’re thinking through similar challenges in your organization, I’d be interested to hear your perspective.
