Retailers already track which channels drive traffic to their site. As AI reshapes discovery, that context may become an increasingly important signal for what should happen when shoppers reach the Transaction Moment™.

When shoppers arrive at the point of purchase, they’re no longer a uniform group. A consumer who compared options in ChatGPT for 20 minutes is in a different headspace than a loyal customer who typed the URL from memory, or a price-driven shopper who tapped through from a comparison ad. They’ve done different amounts of work before arriving. They need different things from the pages where the purchase happens.

Yet many transaction experiences still treat these arrivals the same. We’re calling that mismatch the intent gap.

The mismatch between the context a shopper carries into the transaction and the experience they find when they get there. As AI widens where purchase-ready opinions form, the range of mindsets hitting the same cart, payment, and confirmation pages is growing. What feels helpful to one shopper may feel redundant, distracting, or trust-breaking to another.

Intent Determines Where Journeys Start

Forrester’s vision report, Death Of The Retail Website As We Know It,  provides a useful framework for understanding this fragmentation. Their research, drawn from a February 2026 consumer panel of online adults in the US, UK, and Canada, shows that shopper intent increasingly influences where the digital journey begins.

Different channels attract shoppers with different intent profiles: answer engines for high-consideration research and comparison, brand and retailer sites for loyalty-driven and high-value purchases, and marketplaces like Amazon for commoditized repeat buys.

Forrester’s recommendations focus upstream: assess your answer engine opportunity, build assortments by channel, improve the website experience for both humans and machines, and test continuously. The question is what it means for the transaction itself. 

The Behavioral Evidence

Early behavioral data suggests these journeys are already producing different outcomes. Reported benchmarks from Shopify and Adobe point to rapid growth in AI-referred traffic. Shopify’s Q1 2026 data showed AI-referred sessions growing 8x year-over-year, with AI-referred visitors reportedly converting at nearly 50% higher rates than organic search and carrying 14% higher average order values. Adobe found AI-driven traffic to retail sites grew 393% in the same period, converting 42% more often than non-AI sources with 37% higher revenue per visit. 

At the other end of the spectrum, social traffic often shows materially higher abandonment. How someone arrives is already influencing whether they convert. The same transaction experience, producing very different outcomes depending on who walked through the door and what they’d already decided.

The exact numbers will vary by retailer, category, and source. But the pattern is what matters: the spread between mindsets hitting the transaction is widening, not narrowing. The industry already has a playbook for responding to this kind of signal. It just hasn't carried it far enough.

The Upstream Logic That Stops At The Transaction

The marketing industry already personalizes by traffic source. Dynamic landing pages that adapt headlines, offers, and CTAs by referral context are common practice. HubSpot’s analysis of personalized CTAs is one example of the broader principle: context improves relevance (HubSpot, CTA Personalization Study). But that logic often weakens or disappears once the shopper moves deeper into the transaction.

A retailer running multiple landing page variants by traffic source may still serve a single cart, payment, and confirmation experience to everyone who gets past the product page. The signals that shaped the upstream experience do not always follow the shopper into the moments where the purchase is completed.

If referral context improves relevance upstream, the question is whether it can do the same at the transaction.

Relevance At The Transaction

The transaction is already absorbing strategic work it was never designed to handle. As discovery, comparison, and consideration migrate into AI surfaces, cart, payment, and confirmation pages carry more of the journey than they used to. Our whitepaper, The New Economics of Checkout, made the case that this concentration makes relevance at the point of purchase, including knowing when to show nothing, the differentiator. 

Forrester’s intent research adds a new dimension. If different intents produce different shoppers, then the relevance calculation at the Transaction Moment is not only about who the shopper is. It is also about what they have already done before landing on the retailer’s page.

What’s relevant to a pre-researched shopper at product selection is different from what’s relevant to a loyal repeat buyer at the same stage. Not because they necessarily belong to different customer segments, but because their prior context changes what would be helpful versus what would create friction.

We believe arrival context is becoming an important relevance signal at the point of purchase. As the spread between shopper mindsets widens, it will matter more.

How Arrival Intent Shifts The Relevance Equation

What a shopper has already accomplished before the transaction changes what “relevant” means at each stage.

These are not fixed segments or universal rules. They are testable hypotheses about how arrival context may change what feels relevant across the Transaction Moment.

A relevance engine that accounts for arrival context can make sharper decisions about what to show, when to show it, and when to show nothing. The pattern holds: what someone already accomplished before the transaction is informative about what may be relevant to them during it.

What Retailers Can Do Now

1. Segment transaction performance by arrival source. Most ecommerce teams track checkout conversion as a single number. Split it by traffic source and the first relevance gaps may surface. The first clues are likely already in the data.

2. Identify where expectations break. For each major traffic source, find the transaction stage with the highest drop-off. If AI-referred shoppers abandon at cart review more than direct shoppers, something on that page may contradict the expectations they formed upstream. That turns a broad conversion problem into a more specific, diagnosable one.

3. Carry the landing-page logic through. Test whether the signals that improve conversion upstream, such as referral source, campaign context, and UTM parameters, can improve conversion at the transaction. Start with the highest-volume sources where the intent signal is clearest and measure stage by stage.

4. Audit the transaction against intent profiles. Walk through your own transaction experience from each arrival context. Does the experience make sense for a shopper who already compared five products? For one who has bought from you three times? For one benchmarking you against one-click checkout? Many teams have not looked at the transaction through these lenses.

5. Invest in relevance infrastructure at the point of purchase. Static transaction experiences are poorly suited to respond to different arrival contexts. Rokt’s platform is built to optimize relevance across the Transaction Moment™ using permissioned identity, behavioral, transaction, and contextual signals, including the decision to show nothing. Arrival source and pre-transaction context may become increasingly valuable inputs where they are captured, permissioned, and available. The near-term opportunity is to test which signals improve relevance and outcomes at each stage of the transaction.

Where We Think This Is Heading

The behavioral shift Forrester documents is real and accelerating. Different intent is routing shoppers to different starting points. Those starting points appear to produce measurably different shoppers. And many transaction experiences are not built to account for the difference.

Our whitepaper argued that AI is concentrating brand value at the transaction. Forrester’s research shows that intent is fragmenting where journeys begin. Together, those dynamics raise the stakes on relevance at the point of purchase. The decisions made there carry more weight, and the signals informing them need to reflect more of the context the shopper brings.

This is early. How much value arrival-context signals add will need to be tested. But the direction is clear: as AI changes how shoppers decide, retailers will need to rethink not only where journeys begin, but what happens when shoppers are ready to buy.

The full argument for why the Transaction Moment is absorbing more strategic value is in our whitepaper, The New Economics of Checkout.

Sources

• Forrester, “Death Of The Retail Website As We Know It” (June 2026)
• Forrester, ConsumerVoices MROC Survey, February 2026
• Rokt, “The New Economics of Checkout” whitepaper (2026)
• Shopify, Q1 2026 AI Search Insights
• Adobe Digital Insights, Q1 2026 AI Traffic Analysis
• HubSpot, CTA Personalization Study