Found by the Algorithm, Lost at the Shelf

SPRINGBOARD MANAGEMENT  ·  THE DATA-DRIVEN PULSE  ·  APRIL 2026

Agentic commerce is reshaping how products get discovered, evaluated, and purchased — often before a human shopper ever enters the picture. Our Q1 2026 eConsumer Pulse™ data shows that over half of shoppers across most major CPG categories are already using AI tools when they buy, a shift increasingly understood through data analytics and market research. Most brands are asking the wrong question. It’s not “are we ready for AI?” It’s “can AI find us, trust us, and choose us — and do we even know if it can’t?” a shift increasingly understood through data analytics and market research.

SETTING THE STAGE

The Shelf Moved. Again.

There’s a retail axiom that has held for decades: if you’re not on the shelf, you don’t exist. It started in the grocery aisle. Then it migrated to the digital shelf — page one of search results, the top of a category on Amazon, the hero placement on a retailer’s PDP. Brands adapted. They invested in SEO, content, sponsored placements, and ratings strategies, often supported by data analytics tools and business data analytics capabilities. They learned the rules of digital discovery and competed accordingly.

The shelf has moved again. And this time, the rules are fundamentally different.

Agentic commerce — the model in which AI systems autonomously search, evaluate, and increasingly transact on behalf of consumers — is not a concept being debated in innovation labs. It is being built, deployed, and used right now. Walmart launched Sparky, its agentic shopping assistant, in June 2025. Shopify President Harley Finkelstein called ChatGPT’s Instant Checkout the “new frontier” for online retail. Google launched its Universal Commerce Protocol — an open standard for agentic commerce — co-developed with Walmart, Target, Shopify, Etsy, and Wayfair, and endorsed by more than 20 brands including Chewy, Kroger, Gap, Sephora, Ulta, and The Home Depot. When Visa, Mastercard, and Stripe all sign on to the same protocol, that’s not a press release. That’s infrastructure.

The brands showing up on this new shelf are not winning because of their AI stack. They are winning because their product content, attributes, and proof architecture are clean, structured, and machine-readable. And the brands that are invisible? Most of them don’t know it yet.

What Springboard is seeing: Q1 2026 eConsumer Pulse™ data

Our proprietary quarterly research tracks how consumers use AI tools across 18 eCommerce product categories. The Q1 2026 data makes one thing clear: this is not a niche behavior confined to tech-forward categories. It is mainstream, and it is accelerating across the entire CPG and retail landscape.

Among shoppers actively using AI tools when buying: 72% of alcoholic beverage shoppers, 65% of electronics shoppers, 60% of toys and hobby shoppers, 58% of healthcare and OTC shoppers, 56% of apparel shoppers, 54% of arts and crafts shoppers, and 54% of vitamins and supplements shoppers are all using AI as part of their purchase process. Even categories traditionally considered less digitally driven are crossing the threshold: 51% of non-alcoholic beverage shoppers, 46% of pet food and care shoppers, and 45% of grocery shoppers. In category after category, more than half of the shoppers in your market have already incorporated AI into how they research, evaluate, and buy.

And it is not just research. Shoppers are using AI to complete purchases directly: 21% of toys and hobby shoppers, 20% of beauty and personal care shoppers, 17% of grocery shoppers, and 14% of pet care shoppers report using AI tools to make a purchase or place an order. These numbers will only grow. The consumer has already moved. The question is whether the brands they shop have moved with them.

Figure 1: Springboard eConsumer Pulse™ Q1 2026 — AI usage across 12 CPG & retail categories. Gold labels indicate share completing a purchase via AI. Categories sorted by total AI usage.

THE SPRINGBOARD POV

This Is a Data Problem. Not a Technology Problem.

Here’s what we keep hearing from brands right now: “we’re working on our AI strategy.”, often without a clearly defined data strategy  or the support of data analytics consulting. And when we ask what that means, it usually involves some combination of evaluating platforms, standing up pilots, or exploring what GenAI can do for marketing content.

Those are reasonable things to explore. But they’re not what will determine whether a brand wins or loses on the agentic shelf. What will determine that is something far more foundational: the quality, structure, and credibility of their product data consulting. 

When a consumer asks ChatGPT, Gemini, or Perplexity to recommend a high-protein snack bar, a fragrance-free laundry detergent, or a vet-recommended joint supplement for dogs, the AI is not browsing a digital storefront. It is reading structured attributes, synthesizing reviews, evaluating claims against third-party signals, and returning an answer. If your product attributes are incomplete, inconsistent, or absent, you are not in the conversation. There is no sponsored slot to buy your way back in. There is no end-cap equivalent. You simply do not exist in that moment.

Google Cloud framed it plainly in their CPG agentic commerce guidance: if your product uses sustainable packaging but that information isn’t structured and tagged, an AI agent searching for “verified sustainable packaging” won’t find it. The attribute has to be machine-readable to be machine-recommended. This is not an IT problem. It is a brand and insights problem — one that lives squarely at the intersection of how you understand your consumer and how you present your product to the world.

THE PROBLEM BEHIND THE PROBLEM

You Can’t Optimize What You Can’t See.

Here’s the harder truth beneath the data infrastructure problem: most brands have no visibility into how they’re performing on the agentic shelf at all.

On the traditional digital shelf, performance is measurable. Search rank, PDP conversion, sponsored placement ROI — there are tools, dashboards, and benchmarks. You know when you’re losing ground because you can see it. Agentic commerce, for now, offers no equivalent, increasing the reliance on data reports and digital reports to approximate visibility. There is no “AI shelf share” report in your data analytics tools. When Walmart’s Sparky doesn’t recommend your product, or when a Gemini query surfaces your competitor instead of you, you don’t get a notification. You just don’t get the sale — and you don’t know why.

This is the visibility gap. And it compounds quietly. Every week that passes without understanding how AI systems are representing your brand — or whether they’re representing it at all — is a week your competitor’s structured data is earning trust your brand isn’t building.

The NIQ and Kearney joint analysis on agentic commerce and CPG growth makes a point that should reframe how established brands think about this moment: AI-powered discovery is tilting the competitive landscape toward challenger brands. Why? Because agents don’t have brand bias. They don’t default to the household name because it has a bigger media budget. They surface the most relevant, best-supported answer to the consumer’s stated need. That means an emerging brand with clean, complete, well-reviewed product data can beat an incumbent that’s been on the shelf for 30 years — simply because it shows up better to the machine.

For large CPG brands, that’s a wake-up call. For challenger brands and emerging players, it’s an opening that won’t stay open forever.

WHO’S MOVING — AND WHAT WE CAN LEARN

Early Movers Across Categories

The brands executing well on agentic commerce right now share less in common with each other by category than they do by mindset. They are treating product data as a commercial asset, not a back-office function. A few examples worth watching:

• Walmart (Mass / Grocery): Sparky, launched in June 2025, replaced keyword search with goal-based shopping — customers state an intent like “I’m hosting a cookout” and Sparky plans, reasons, and fulfills across Walmart’s entire retail stack. In January 2026, Walmart deepened this through a Google Gemini integration via the Universal Commerce Protocol, making its full assortment discoverable through conversational AI. The bet: be the easiest brand for AI to recommend.

• Gap Inc. (Fashion / Apparel): Made its products shoppable through Google Gemini via UCP, enabling seamless checkout across AI-native environments. The framing from their CTO: “We are not pursuing AI for novelty.” The products that win in agentic fashion discovery will be those with the richest attribute sets — fit, material, occasion, sustainability credentials — all structured and machine-readable.

• Macy’s (Department / Beauty): Launched “Ask Macy’s,” an agentic AI experience built around curated discovery rather than search. “It’s not about search,” said Macy’s chief customer officer at ShopTalk. “It’s about curated discovery — what they need and what they want.” The implication for brands selling through Macy’s: if your products aren’t surfaced by that agent, you’re losing consideration before the shopper ever browses.

• Petco (Pet): Among the first pet retailers collaborating with Google on Direct Offers, the next-generation agentic discovery layer. In a category driven by trust, health claims, and vet recommendations — exactly the kind of structured, verifiable attributes that AI agents can evaluate — early movers will define the category standard.

• Challenger Brands (All Categories): The NIQ-Kearney research makes a compelling case that this moment disproportionately benefits emerging players. Agentic commerce democratizes discovery. A well-positioned challenger with clean data, strong reviews, and clearly articulated attributes can appear alongside — or ahead of — incumbents that haven’t invested in their content infrastructure. That is a genuine strategic opportunity for brands willing to move now, all driven by strong business data and efficient business operations.

What unites these examples is not the sophistication of their AI implementation. It is the quality of the foundation underneath it — the product truth, the review architecture, the attribute completeness that makes a brand legible to a machine making decisions at a speed no human analyst can match.

THE MARKETING SHIFT NOBODY BUDGETED FOR

Agentic Commerce Isn’t Just Changing Discovery. It’s Changing Where the Money Goes.

Every major shift in how consumers shop has eventually triggered a corresponding shift in how brands spend their marketing dollars, in strategy consulting, business strategy consulting, and advisory services. When the digital shelf emerged, budgets migrated toward search, sponsored placements, and retail media. Agentic commerce is the next inflection point — and it will require a rethink of marketing allocation that most brands haven’t started yet.

The core tension is this: you cannot buy your way onto the agentic shelf the way you bought your way onto page one. Traditional retail media — sponsored placements, display, programmatic — is built on the premise that attention is purchasable and eyeballs can be directed. Agents don’t have eyeballs. Perplexity said it plainly when Amazon sued them over their shopping agent: AI agents “don’t have eyeballs to see the pervasive advertising Amazon bombards its users with and cannot be upsold to buy more products.” That’s not a product critique. It’s a description of how the economics of discovery are changing.

Kantar Retail IQ frames the shift clearly: in an agentic world, retail media moves from persuasion to negotiation. Agents compare total value in real time — price, availability, loyalty benefits, delivery speed, product attributes — and surface the best answer. That changes the ROI conversation for every dollar a brand puts into retail media. Impressions and CPMs weaken as metrics. Structured data quality, review velocity, attribute completeness, and content credibility become the new media currency. The brands that invest in those foundations now are building the kind of durable algorithmic trust that paid media can’t replicate and competitors can’t easily buy out from under you.

What the budget reallocation actually looks like

This isn’t an argument for abandoning retail media or pulling back from paid channels. It’s an argument for adding a new line item to the marketing budget conversation — one that most brand P&Ls don’t have a category for yet. Think of it in three buckets:

• Content infrastructure investment. Cleaning, enriching, and structuring product data across every channel where an AI agent might encounter your brand. This is not glamorous work. It is also not optional. Brands that treat product content as a marketing asset — and budget for it accordingly — will compound returns as agentic commerce scales. Those that treat it as a back-office function will find themselves invisible on a shelf they didn’t know existed.

• Review and reputation architecture. Reviews are the proof layer that AI agents trust most. Volume, recency, specificity, and sentiment are all signals that influence how agents rank and recommend products. A brand strategy that actively cultivates high-quality, keyword-rich, attribute-specific reviews is building a marketing asset that works 24 hours a day inside every AI interface where a consumer might ask about your category. That’s media reach without a media buy.

• Generative Engine Optimization (GEO). Just as SEO required brands to rethink how they wrote and structured content for search engines, GEO requires rethinking how content is written and structured for AI systems. This includes everything from how product descriptions are written to how brand messaging is framed on owned channels that LLMs crawl. Brands that invest in GEO now are building discoverability that compounds over time — the same way strong SEO built organic search equity for years.

The challenger brand advantage — for real this time

Here is where the monetization shift becomes genuinely disruptive for the competitive landscape. A challenger brand with a $2M annual marketing budget that redirects a meaningful portion toward content infrastructure, review cultivation, and GEO can outperform a $50M incumbent that is still allocating the majority of its spend toward traditional retail media and brand awareness. That’s not hyperbole — it’s the structural implication of a world where the buyer is an algorithm that doesn’t respond to creativity, doesn’t see an end-cap, and can’t be retargeted.

But there’s a critical caveat that applies to challenger brands and incumbents alike: agentic commerce can surface your brand — it cannot fulfill it. An AI agent that recommends a product not in stock at the consumer’s preferred retailer, or without warehouse presence for two-day delivery, will simply substitute. It doesn’t wait. It finds the next best answer. Distribution remains the bridge between discovery and conversion, and brands that invest in GEO and content infrastructure without the retail relationships and supply chain readiness to back it up are building a funnel with a hole at the bottom. The invisible shelf and the physical shelf have to work in concert. For challenger brands especially, this means that winning on data is a necessary condition for agentic commerce success — but it is not sufficient. The harder, longer work of earning and expanding distribution doesn’t go away. It becomes more urgent.

For large CPG and retail brands, this is a strategic reallocation conversation that needs to happen at the CFO and CMO level, not just in the digital marketing team. The question isn’t how to add GEO as a line item. It’s how to evolve the entire marketing investment thesis for a world where a growing share of purchase decisions will be made — or heavily pre-filtered — by AI before a human ever weighs in.

THE PRACTICAL PATH FORWARD

Four Questions Every CPG and Retail Brand Should Be Asking Right Now

You don’t need a full agentic commerce strategy to start moving. You need clarity on where you stand. These four questions cut through the noise:

• Is your product content machine-readable? This means more than having a PDP. It means structured attributes, verified claims, complete category taxonomy, and rich descriptive data that an AI can evaluate against a consumer’s stated need. If you haven’t audited your content against the standard of what an AI agent needs to surface your product — not what a human browsing needs — you don’t know where your gaps are.

• Do you have a signal on how AI systems represent your brand? Most brands don’t. That’s the visibility gap. Start by testing. Query your own products across ChatGPT, Gemini, and Perplexity across a range of consumer need states. What comes up? What doesn’t? Where do competitors appear instead? This is a manual exercise today, based on the effective use of data analytics tools and structured data reports, but it is one of the most revealing brand audits you can run — and it costs nothing but time.

• Are your reviews earning trust from algorithms, not just humans? Salsify’s 2026 consumer research found that 47% of shoppers cite positive ratings and reviews as a top reason for trusting brands. AI agents read that same signal. Review qualitative and quantitative research, recency, specificity, and volume are all inputs into how AI systems evaluate product credibility. If your review strategy is designed to move star ratings, it’s not fully designed for agentic commerce.

• Do you have the tools and capability in-house to assess and act on any of this? This is the question most brands skip — and the most important one. Auditing your agentic readiness, identifying content gaps, building a GEO strategy, and monitoring how AI systems represent your brand requires a specific combination of shopper insights expertise, data infrastructure knowledge that may require training in data analytics, management workshops, and AI fluency that most brand teams haven’t yet developed internally. Knowing what to look for is half the battle. Having the capability to act on it is the other half. If the honest answer is “not yet,” that’s exactly where Springboard can help — from initial audit through to execution and team upskilling.

Our Q1 2026 eConsumer Pulse™ data shows that in most CPG categories, more than half of shoppers are already using AI tools somewhere in their purchase journey. That share will only grow. The brands that answer these four questions with clarity and confidence today are building the agentic shelf presence that will be very hard to replicate once the window closes. This is not a 2027 problem. The infrastructure is being built today, the consumer behavior is shifting now, and the early movers are already pulling ahead.

Sources: BoF–McKinsey State of Fashion 2026  ·  Kantar Retail IQ Agentic Commerce Predictions (2026)  ·  Google Cloud CPG Agentic Commerce (2026)  ·  NIJ–Kearney “New Growth Frontier” (2026)  ·  Salsify 2026 Consumer Research Report  ·  Dataiku Retail & CPG Trends 2026  ·  Walmart / Google UCP Announcement (Jan 2026)  ·  eMarketer Agentic Commerce & CX (2026)  ·  Mirakl Retail Media Trends 2026  ·  Kellogg School of Management: Walmart’s Sparky (2026)

springboardmgt.com  ·  © 2026 Springboard Management, Inc. All rights reserved.