The Answer Is the New Shelf: What CMOs Should Learn from the Stroller and Car Seat Category

Every category has a moment when a new distribution channel stops being optional. For retail, it was the shift from catalog to storefront to e-commerce. For media, it was the shift from broadcast to cable to streaming. Each transition looked, in its early years, like a niche behavior confined to a small slice of early adopters. Each one turned out to be the whole game within a decade.

A new study out of 5W AI Communications gives CMOs in consumer categories a concrete, quantified look at where we are in the current transition: the shift of buyer research from search engines and store shelves to generative AI answer engines. The Strollers & Car Seats AI Visibility Index 2026 analyzed 2,400 responses across ChatGPT, Claude, Gemini, Perplexity, and Google AI Overviews to determine which brands these tools recommend when parents ask about strollers and car seats. The findings are specific to one category, but the mechanics they expose apply well beyond it, and they should reshape how marketing leaders think about brand equity, retail distribution, and reputational risk.

The Headline Number Isn’t the Real Story

UPPAbaby captured 19.8% of citations, the largest share of any brand tracked, with Chicco and Graco rounding out the top three at a combined 47.8%. That is a useful data point for anyone in the category, but it is not the finding that should reorient a CMO’s thinking. The more instructive number is Evenflo’s: 3.6%, a share well below what its retail distribution at Walmart, Target, and Amazon would predict, correlating with a 2020 safety investigation into the brand’s booster seats that the study found AI engines are still surfacing today.

Sit with the shape of that gap for a moment. Retail distribution, unit sales, shelf space, none of the traditional inputs to market share explain it. What explains it is a specific, well-documented reputational event from six years ago that has apparently not been displaced in the AI-generated answer, even as three other manufacturers named in the same industry-wide Congressional inquiry show no comparable penalty. This is a case where a brand’s real-world retail performance and its AI-era answer-share have detached from each other, in a way that traditional market research, focused on sales, share of shelf, and social sentiment, would not have caught.

That detachment is the pattern CMOs need to internalize. Citation share and market share are correlated, but they are not the same metric, and the gap between them can widen quietly, without a single quarterly sales report reflecting it, until the AI-native buying behavior it predicts starts to show up in the numbers a year or two later.

Four Brands, One Product Each

The second finding worth a CMO’s attention is about where citation share actually concentrates. Four of the top five brands in the study, UPPAbaby, Chicco, Doona, and Nuna, carry the majority of their AI citation share from one or two flagship products, not an even distribution across a broad catalog. Doona’s entire 9.7% share rests on a single hybrid car-seat-and-stroller product with no direct structural competitor. Chicco’s 15.4% is disproportionately carried by one infant seat model, whose crash-test performance is independently documented by Consumer Reports.

This should reframe how marketing leaders think about portfolio strategy in the AI-search era. The traditional playbook rewards a broad, well-marketed catalog: multiple price points, multiple styles, something for every segment. That playbook still matters for retail merchandising and brand breadth. But it is not what wins the AI-generated answer. What wins there is depth: one product with a dense, independently authored body of reviews, comparisons, and safety documentation that an AI retrieval system can draw on with confidence. A flagship product with deep documentation appears to outperform a broad catalog with shallow documentation, category by category, brand by brand, in this data.

For a CMO managing a multi-SKU portfolio, that suggests a genuinely different resource allocation question: which one or two products in the line deserve the deepest possible independent documentation, review density, and structured data, because that is the product most likely to become the entity an AI engine reaches for when a customer asks a category-level question.

The Retrieval Logic Behind Both Findings

Both patterns, Evenflo’s suppressed share and the flagship-product effect, trace back to the same underlying mechanism, and understanding that mechanism is more useful to a CMO than memorizing either finding in isolation. Generative AI engines do not evaluate brands the way a human researcher does, weighing overall impression and recent sentiment. They retrieve specific, well-documented, attributable content and construct an answer from it. A dated crash-test score is retrievable. A specific investigative report with internal documents is retrievable. Vague brand marketing language, spread evenly across a wide catalog, is comparatively low-value to a retrieval system, because it does not resolve a specific factual question.

That single mechanism explains why a flagship product with dense, independent documentation outperforms a broad catalog with thin documentation. It also explains why an old, specific, well-sourced investigation can outperform a brand’s current, general safety messaging in the same retrieval pool. In both cases, specificity wins. The system rewards content that answers a precise question with a precise, attributable fact, regardless of whether that content favors the brand or works against it.

What This Changes About the CMO’s Job

The practical implication is that AI citation share deserves its own line in the marketing dashboard, tracked with the same discipline as share of voice, share of shelf, and social sentiment, and audited on a recurring basis by actually asking the major AI engines the buyer-intent questions a customer would ask. It also means the content investment case for a brand’s most important product needs to include a category that most marketing budgets do not currently have a clean home for: independently verifiable, structured, dated documentation, crash-test results, certifications, vehicle-fit data, that exists specifically to be retrieved and cited by a machine, not just read by a person.

And it means reputation management needs a genuinely new time horizon. The 5W findings suggest that a well-documented negative event does not fade from an AI-generated answer the way it fades from a search results page or a news cycle. If a brand carries a documented safety event, a recall, an investigation, a public finding, silence is not a strategy. The gap only closes if the brand produces content at least as specific and well-documented as the negative event itself.

The category here is strollers and car seats, but the lesson is not. Any consumer brand competing in a market where buyers now bring their questions to an AI engine before they bring them to a search bar or a store aisle should read this study as a preview of its own dashboard, whether or not that gap has shown up in quarterly sales yet. The answer is the new shelf. The brands that treat it that way, now, will be the ones citation share favors a year from now.

Sources

• 5W AI Communications — Strollers & Car Seats AI Visibility Index 2026

• ProPublica — Evenflo, Maker of the “Big Kid” Booster Seat, Put Profits Over Child Safety (Feb. 6, 2020)

• House Oversight Committee — Investigation into Booster Seat Safety

• Consumer Reports — Chicco KeyFit 30 review and crash-test score

• 5W Parenting Apps, Gear & Retail practice

Ronn Torossian is Founder and Chairman of 5W Public Relations, the AI Communications Firm, and has advised Fortune 500 companies on crisis communications and regulatory reputation for more than two decades.