Invisible negotiations shaping commerce
Your brand’s AI agent is in a silent negotiation. A consumer’s agent has just asked for “the best detergent.” In milliseconds, your agent presents the facts: lowest unit price, certified sustainable, fastest shipping. Another brand’s agent counters with a coupon.
A third highlights superior stain removal. The consumer never sees this exchange, but the outcome decides who wins the sale. This is the new shelf: invisible to people but imperative for growth.
The business problem: invisibility and unpreparedness
For decades, consumer decision-making followed a predictable pattern: marketing built awareness, retailers controlled availability, and loyalty programs kept shoppers in orbit. In the agent-mediated world—agentic commerce—that logic collapses. Discovery no longer depends on a store shelf or a search ranking; it depends on what data an algorithm can see, read, and trust.
Industry research from 2025 shows only 37% of brands conduct a monthly audit of how AI assistants describe them, while 25% do so just once a year. Even more concerning: no standard or automation exists for these audits, revealing a deep misunderstanding of readiness itself. If consumers turn to commerce agents that only see spec sheets, the $43 billion U.S. consumer goods companies spend on advertising each year could vanish into a single line on a spreadsheet.
The opportunity: a once-in-a-generation reset
Agentic commerce upends the retailer-dominated model of the past century. Brands that once rented shelf space can now own their visibility if they build for machines, not people.
But recent research shows how far brands still have to go. Only 37% of brands audit how AI assistants describe them monthly, and 4% of brands less than once a year. Only 33% say their product data is “very consistent” across channels. Only 36% describe their data as “fully structured and machine-readable.”
In other words, the shelf is collapsing, but the window to rebuild it is open. Most consumer product companies think they’re fully mapped, but they don’t know what they don’t know.
Mobile changed how agents change who
At first glance, agent-to-agent commerce might seem like another passing hype cycle. Yet much like how mobile technology redefined how people connect, agents have the potential to redefine who is doing the connecting.
Campaigns once driven by imagery and emotion must now be underpinned by structured facts—because algorithms don’t care for clever taglines. They parse clarity, comparability, and proof.
Most industry leaders believe in this future but overestimate their readiness for it. Researchindicates that 64% claim to have a company-wide strategy for influencing how AI tools describe their products, yet many of these strategies remain largely theoretical. This strategy belongs in the boardroom, requiring executive leadership to translate business goals into data structures that agents can read and rank.
Five gaps that define risk
Industry research identifies five readiness gaps that define the next competitive fault line:
The audit gap – Only 37% audit monthly; 4% audit less than once a year. Most lack automation or defined processes. Auditing isn’t checking a few SKUs manually; it’s a continuous process that tracks how AI assistants describe thousands of products in real time.
The control gap – Retailers dominate visibility: 68% of brands rely on retailer sites versus 52% on brand.com. Even “brand-managed” retail pages live inside retailer ecosystems, meaning brands still don’t control what algorithms read first.
The data gap – Only 33% report very consistent product data across channels. Inconsistency is the new invisibility. If your facts conflict across feeds, agents will choose the competitor with cleaner data.
The accountability gap – Only 39% have a dedicated AI discovery lead, while the rest divide responsibility across digital and brand teams. Ownership remains fragmented; brands are treating an existential issue as a task, not a mandate.
The perception gap – 63% worry competitors will appear more prominently in AI results, but fewer than a third act with urgency. Brands know what’s coming; they just assume it’s a tomorrow problem.
The agent era, in three acts
The rise of agent commerce won’t happen all at once. It will unfold in phases, each one reshaping brand strategy.
Today: AI as a shortcut. Consumers already use AI to search the web for recommendations.ChatGPT sees more than 2.5 billion requests daily, which means product data is already being scraped, whether brands are prepared for it or not.
Near future: AI as a storefront. As consumers offload low-consideration decisions (nearly 77% of U.S. consumers already use AI to make faster decisions), they stop scrolling traditional search engines and marketplaces, going straight to AI tools for transactions. That means fewer human touchpoints, fewer banner ads, and a rising premium on data trust signals. The brands that win will build trust signals—like proof of authenticity, sustainability, and performance—into the data itself.
Far future: Agent-to-agent commerce. Personal devices will default to AI agents, and brands must meet them with commerce agents of their own. Bots will negotiate directly on price, availability, delivery speed, and trust signals. To compete, brands will need autonomous agents, codified negotiation rules, and embedded brand values. Within 24 months, agents will be as important as influencers in shopping choices.
CX meets AX: two shopping missions, two playbooks.
The future splits into two: agents handle convenience missions, while humans still drive experience missions. And because many consumers will still double-check results or distrust a bot’s judgment, customer experience storytelling and human-designed experiences remain essential for trust.
Agents on convenience missions will be like digital concierges, checking stock, comparing prices, and replenishing essentials. The transaction is fast, functional, and invisible. Yet “convenience” can live in the luxury space too: think of a high-net-worth individual who must have that handbag and wants it delivered within hours.
Experience missions are driven by inspiration, aspiration, or curiosity—like browsing handbags on vacation, or rediscovering joy in a weekend grocery or homewares shop. Here, customers seek engagement and storytelling. Even routine categories can become experience-led when time and attention allow.
The future isn’t CX versus agent experience (AX)—it’s CX layered with AX, shifting dynamically with consumer intent. Machines may make the shortlist, but humans still need a reason to care.
Beyond visibility: the data dividend
The prize for machine-readable data goes beyond visibility. Agent-driven discovery gives brands better demand signals, lower customer acquisition costs, and richer feedback. Imagine not just tracking what sold but knowing why an agent recommended it—fragrance-free, certified sustainable, or diet-friendly. That intelligence powers the next level of personalisation and product innovation.
For challenger brands, this may be the best opening in a generation. Research already shows that agents don’t favour history or scale but structured clarity.
Beware the new middleman
But the fall of one gatekeeper risks the rise of another. Aggregators are already testing platform commerce agents that scrape brand data and insert themselves between the consumer and product.
Agents already influence what shows up in social feeds, invisibly curating exposure to brands beyond paid ads—a preview of how algorithmic mediation is expanding from content to commerce. Without an AX strategy, companies may simply swap one middleman for another.
Flight booking is a cautionary tale: by the late 2000s, travel aggregators like Kayak and Skyscanner were pulling in millions of users monthly by scraping airline sites. Airlines were then forced to pay hefty referral fees to reach customers they already had. The same dynamic could hit consumer brands: either be invisible to commerce agents or pay margin-killing tolls to intermediaries.
What leaders should know: the new rules of AX
Here are the concepts that matter now:
AX isn’t SEO. Treating agent experience (AX) like a checklist is the fastest way to lose. It belongs in the boardroom because it affects the entire operating model. Product teams must design with codifiable proof points, like sustainability certifications, provenance, and performance claims.
Training data is the new media buy. The real fight is getting your facts and values into the sources AI agents draw from.
Your brand isn’t a place; it’s a feed. Think of it as a stream of structured facts flowing into the machines that make choices for your customers.
In the future, brands will be judged by their agents. Consumers may never see most of your ads or promotions. What they see is what their AI agent decides.
The takeaway
The shelf is no longer rented space in a retailer’s aisle but a contest of clarity, structure, and trust. Done right, AX can be transformative: a challenger brand coding every claim to leapfrog incumbents, or a luxury house proving not just that a bag is leather, but that it’s ethically sourced and guaranteed. This will be winner-takes-all territory—whoever becomes the default for “best detergent” wins everything.
