Commerce leaders have spent the last decade optimizing a familiar loop: earn attention, drive a click, convert a session, repeat. Even when the channels changed—search to social, desktop to mobile—the underlying control points stayed legible. You could see where demand came from, you could buy more of it, and you could instrument the journey well enough to defend your decisions.
What’s emerging now is a different kind of shift: discovery is detaching from the open web and reassembling inside systems that don’t behave like channels. They behave like intermediaries—AI shopping assistants, digital wallets and platform-governed layers that decide what gets surfaced, how it’s framed and whether it’s affordable, in the moment.
That’s not new marketing. It’s a reallocation of leverage.
AI Assistants as the New Front Door
Start with the front door. AI shopping assistants aren’t arriving as a novelty feature; they’re arriving as a new mediation layer between intent and inventory. The key change isn’t that people will ask an assistant for a recommendation. People have always asked for recommendations.
The change is that the assistant becomes the interface where shortlisting happens, and shortlisting is where brands are made and unmade. When OpenAI and Perplexity push deeper into shopping behaviors, they’re not competing with one retailer; they’re competing with the act of browsing itself. You don’t need a shopper to visit ten sites if the assistant can compress the evaluation into one answer.
That compression is both the strategic threat and the opportunity: it reduces customer effort while raising the stakes on LLM brand visibility inside the systems that generate the answer.
One of the more revealing signals isn’t a product launch at all; it’s the measurement of which retailers and brands appear most often in AI shopping mentions during the holidays. Ulta, Best Buy and Adidas showing up prominently isn’t just a leaderboard; it’s a leading indicator of a new kind of shelf space.
In the old model, shelf space was search rank, marketplace placement, or paid impression share. In this model, shelf space is answer inclusion. That’s a different contest. It rewards structured product information, consistent entity signals, and content that machines can reliably interpret. It also introduces a quiet new form of competitive advantage: the brands that can be correctly represented by automated systems win attention without necessarily winning the click.
Peak Commerce Becomes a Systems Test
Zoom out to the peak shopping moment, because Black Friday is where behavior changes show up first. When record online spending and retail performance narratives focus on winners and losers, the useful signal isn’t the numbers in the headlines. It’s what those peak moments reveal about the new system: consumers are shopping under constraint, decision-making is getting more assisted and conversion is increasingly shaped by the layers around checkout, including payments, eligibility, friction and trust.
Traffic still matters, but it’s no longer the bottleneck. Orchestration is. If your offer, your inventory and your payment options don’t align under load, you don’t just lose a sale; you lose future inclusion when assistants and platforms learn which experiences disappoint.
This is where payments stop being a back-office concern and become a discovery amplifier. When an executive from a buy-now-pay-later (BNPL) provider argues that AI will redefine shopping and payments, it’s easy to read that as industry self-interest. But pair it with regulator-grade reporting and you get a clearer picture: BNPL as commerce infrastructure has reached a scale where it shapes behavior, not just conversion.
As more shoppers use installment options to manage timing, essentials, and cash flow, BNPL starts to function less like impulse fuel and more like a budgeting tool. That shift changes the operating assumptions for merchandising, promo planning and returns. If affordability becomes part of the product experience, then the shopping decision isn’t just “Do I want this?” It’s “Can the system make this feel safe to buy right now?”
Wallets, Affordability, and the Hidden Conversion Layer
Wallets intensify that dynamic. The U.S. payments landscape underlines a structural point that merchants often underrate: wallets are becoming the wrapper that carries not just payment credentials, but identity, authentication, and increasingly the friction profile of the checkout itself.
As wallet-first checkout becomes the default, optimization shifts away from your handcrafted checkout and toward how well you perform inside wallet-native flows — authorization rates, tokenization readiness, fraud posture and the consistency of the data you present to upstream systems. Conversion becomes partly a systems-integration outcome, not solely a UX outcome.
At the same time, the monetization tactics that powered growth in the last era are coming under governance pressure. A New York law targeting personalized pricing is a reminder that personalization is no longer a purely technical or marketing decision. Surveillance pricing regulation is pushing individualized offers and dynamic pricing into the realm of compliance and trust. What once looked like quiet optimization now carries disclosure expectations and potential scrutiny.
When that collides with the reality of returns economics repricing—fees added, windows tightened, policies clarified—the direction is consistent: commerce is moving from persuasion-first to accountability-first. Not because brands want to be restrained, but because the environment is forcing clearer rules on practices that used to be invisible.
The Measurement Gap Nobody Can Ignore
Put these signals together and you get the throughline: discovery is becoming system-mediated, and system-mediated commerce demands a different executive posture. You can’t treat AI assistants as a content trend, wallets as a payments detail, BNPL as a conversion widget, and personalization as an A/B testing playground. Those are now connected layers in the same stack, and they influence what gets recommended, what gets trusted, what gets financed, and what gets repeated.
The commerce leader’s job shifts from optimizing the funnel to securing a position in the intermediated stack. That means making your brand and catalog machine-readable, making your checkout resilient to wallet defaults, designing affordability as a controlled experience rather than an ad hoc discount, and creating governance that prevents growth tactics from turning into regulatory or trust debt.
The uncomfortable part is measurement. As discovery moves into assistants and wallets, the old instrumentation falls behind. Dashboards may show stable or declining SEO, noisy paid media performance, and uneven conversion, while the real influence shifts upstream to systems deciding what shoppers see before they ever land on your site.
This commerce measurement gap is the new blindness risk. Brands that treat answer inclusion, wallet performance, and affordability uptake as first-class signals will see the shift earlier—and move faster—than those still waiting for attribution to catch up.
The Net Effect
For CEOs
The leverage point is moving from channel optimization to system position. Your growth strategy has to assume intermediaries will increasingly decide what gets surfaced and which experiences earn repeat inclusion, which means investment priorities shift toward integration strength, data clarity, and governance—because those are the new prerequisites for demand.
For CMOs
Brand visibility is becoming partly machine-mediated, which changes the playbook for content, offers, and measurement. You’ll need to treat AI exposure and assistant-driven influence as real top-of-funnel surfaces, while aligning affordability narratives and promo strategy to consumer constraint without training the market to wait for discounts.
For CDOs
You’re now responsible for making the business legible to automated systems: structured product data, clean identity and authentication flows through wallets, reliable eligibility and pricing logic, and instrumentation that can detect influence even when the click never happens. The data model becomes a growth model.
References
OpenAI and Perplexity are launching AI shopping assistants, but competing startups aren’t sweating it — TechCrunch
https://techcrunch.com/2025/11/25/openai-and-perplexity-are-launching-ai-shopping-assistants-but-competing-startups-arent-sweating-it/
Ulta, Best Buy and Adidas dominate AI holiday shopping mentions — Modern Retail
https://www.modernretail.co/technology/ulta-best-buy-adidas-dominate-ai-holiday-shopping-mentions/
AI set to redefine shopping and payments, Affirm CEO says — Reuters
https://www.reuters.com/business/retail-consumer/ai-set-redefine-shopping-payments-affirm-ceo-says-2025-11-19/
Winners and losers of Black Friday 2025 — Retail Dive
https://www.retaildive.com/news/winners-losers-black-friday-2025/806610/
New York state law takes aim at personalized pricing — TechCrunch
https://techcrunch.com/2025/11/29/new-york-state-law-takes-aim-at-personalized-pricing/