How AI Search Is Changing the Way Consumers Discover Brands

Industry & Competitive Context

Between 2024 and 2026, the mechanism by which consumers find out that a brand or product exists underwent a structural change. For nearly three decades, the search engine results page a ranked list of ten blue links was the dominant interface between a consumer’s question and a brand’s answer. That interface is now competing with a new one: the AI-generated answer, delivered directly inside Google’s AI Overviews, OpenAI’s ChatGPT, Perplexity, and Microsoft Copilot, among others.

Google began rolling out AI Overviews broadly in May 2024 and by 2026 these AI-generated summaries were appearing in a large and growing share of Google searches, with industry tracking cited by McKinsey estimating that roughly half of Google searches already produced an AI summary, a share projected to exceed 75 percent by 2028. Separately, research firm Gartner forecast in 2024 that traditional search engine volume would fall 25 percent by 2026 as consumers shifted queries to AI chatbots and virtual agents a widely cited projection that several industry publications have since revisited and, in some cases, questioned on methodological grounds, underscoring that it remains a forecast rather than a settled fact.

What is better documented is the traffic that AI platforms are already sending to brand and retail websites. Adobe Analytics, drawing on more than one trillion visits to U.S. retail sites, reported that traffic to those sites from generative AI sources rose 1,200 percent in February 2025 compared with July 2024, with growth roughly doubling every two months since September 2024. During the November–December 2024 holiday period, AI-referred traffic to U.S. retail sites was up 1,300 percent year over year, and on Cyber Monday 2024 it was up 1,950 percent year over year. Adobe’s accompanying survey of more than 5,000 U.S. consumers found that 39 percent had already used generative AI tools for online shopping and 53 percent planned to do so within the year; of those who had used it, 92 percent said it improved their shopping experience and 87 percent said they were more likely to use it for larger or more complex purchases.

Competitively, this has reframed who brands are actually competing against for visibility. BCG’s 2024 analysis of the shift noted that where a traditional search results page might surface dozens of links for a consumer to sift through, a large language model’s answer may surface only two or three products or brands directly in its response compressing what used to be a multi-click, multi-session research process (BCG described consumers researching a major purchase “possibly over weeks”) into a single, synthesized answer. BCG also pointed to early monetization moves by AI platforms: Microsoft Copilot delivering AI-generated answers alongside display advertising similar to conventional search, and Perplexity backed by investors including Nvidia and Jeff Bezos opening the door to sponsors influencing suggested follow-up questions, while OpenAI’s leadership, including CEO Sam Altman, has stated a preference for subscription revenue over advertising without ruling advertising out entirely.

Brand Situation Prior to the Shift

Before this period, brand discovery operated largely on the logic of traditional search engine optimization (SEO): visibility was a function of a brand’s own website ranking on page one for relevant keywords, supplemented by paid search and, increasingly, social and marketplace discovery. Brand marketing teams measured success substantially through organic rankings, click-through rate, and owned-site traffic.

That logic no longer fully applies inside an AI-generated answer. McKinsey’s analysis of Google AI Overview source citations found that a brand’s own website accounts for only 5 to 10 percent of the sources an AI search engine actually references when constructing an answer about that brand. In sectors such as consumer-packaged goods and financial services, McKinsey found that more than 65 percent of the sources cited in AI search answers were third-party publishers, user-generated content, and affiliate sites rather than brand-owned properties. In other words, the party answering a consumer’s question about a brand, inside an AI engine, is frequently not the brand itself — a material departure from the search era, where a well-optimized brand website could at least compete directly for the top organic position.

This gap was compounded by low organizational readiness. A McKinsey CMO survey conducted in September 2025, covering roughly 30 Fortune 500 consumer brand marketing leaders, found that only 16 percent of brands were systematically tracking their performance inside AI search engines at all meaning the large majority of major consumer brands had, at that point, no structured visibility into how often, how accurately, or how favorably they were being represented in AI-generated answers.

Strategic Objective

The strategic problem facing brand marketers, as this body of public research frames it, is not simply “optimize for a new channel.” It is to secure favorable, accurate representation inside a discovery layer the brand does not own, cannot fully measure with legacy tools, and in which the pool of third-party sources (reviews, forums, publisher content, affiliate sites) rather than brand-controlled messaging is doing a disproportionate share of the talking.

McKinsey frames the resulting objective for brands as closing a readiness gap before the cost of inaction compounds: its analysis warns that brands unprepared for this shift could see traditional search traffic decline by 20 to 50 percent, and that even brands actively practicing generative-engine optimization (GEO) may currently lag the performance of well-established SEO programs by 20 to 50 percent meaning the discipline is still immature even among adopters. The explicit strategic objective documented across these sources is therefore twofold: (1) build visibility and measurement into AI-generated answers specifically, rather than only traditional search results, and (2) strengthen the quality, volume, and accuracy of third-party and owned content that AI engines draw upon, since that content not the brand’s own site is what AI engines cite most.

Response Architecture & Execution

Because this is an industry-wide behavioral shift rather than a single branded campaign, the documented “architecture” is a set of parallel responses by consulting firms, analytics providers, and AI platforms themselves, rather than one company’s execution plan.

On the measurement side, Adobe built out AI-referral tracking inside Adobe Analytics specifically to let retailers isolate and quantify traffic arriving from generative AI sources the data source behind its 1,200 percent and 1,300 percent figures allowing retail marketing teams to treat AI referrals as a distinct, trackable channel rather than folding it into generic “other” or “direct” traffic buckets, which had previously obscured its growth.

On the strategy side, McKinsey and BCG have each published frameworks urging brands to treat AI search as a discovery channel requiring its own content and measurement discipline, often described in the broader industry as generative engine optimization (GEO) or answer engine optimization (AEO) the practice of structuring content so that it is more likely to be retrieved, cited, and accurately summarized by large language models, as distinct from traditional keyword-based SEO. BCG’s research specifically flagged that because AI answers surface only a handful of options rather than a full results page, the cost of being excluded from that short list is proportionally higher than being ranked lower on a traditional search page.

A separate industry analysis the Moloco-BCG “AI Disruption Index,” based on a survey of 283 senior marketing leaders across 15 industry verticals and five regions, in-depth interviews with 15 C-suite and VP-level executives, and performance analysis of more than 3,000 mobile apps segmented industries by how exposed they are to this disruption, placing travel, news, education, retail, and auto marketplaces in a “Breached” category of high exposure, and financial services, fintech, media and streaming, and on-demand services in a “Secured” category of comparatively lower exposure. That report’s recommended executive priorities, as reported, were investing in owned digital experiences such as mobile apps, building and activating first-party data, and rebalancing marketing spend toward in-app marketing, CRM, and retail media channels less dependent on being discovered through an open AI search query.

Positioning & Consumer Insight

The consumer insight underlying this shift, as documented in McKinsey’s AI Discovery Survey (fielded in August 2025 across a U.S. panel of 1,927 respondents), is that AI search has become a trusted and preferred research tool for a meaningful share of consumers, not a novelty. The survey found that 44 percent of AI search users already call it their primary and preferred source of purchase-related insight ahead of traditional search engines (31 percent), retailer or brand websites (9 percent), and review sites (6 percent). It also found that more than 70 percent of AI search users ask “top-of-funnel” questions learning about a category or brand for the first time rather than narrow, bottom-of-funnel questions, indicating that AI search is increasingly where brand awareness itself is formed, not merely where a decision already in progress is finalized. Adoption was not confined to younger consumers: McKinsey’s survey found that a majority of baby boomers were already adopting AI-powered search as well, and that 40 to 55 percent of consumers across sectors including consumer electronics, grocery, travel, wellness, apparel, beauty, and financial services were using AI-based search specifically for purchase decisions.

Adobe’s consumer survey adds texture to what shoppers use AI search for: research (55 percent of generative-AI shoppers), product recommendations (47 percent), finding deals (43 percent), gift ideas (35 percent), finding unique products (35 percent), and building shopping lists (33 percent). In travel, the Moloco-BCG index and Adobe both point to high AI usage: Adobe found 29 percent of surveyed consumers had used generative AI for travel planning, most commonly for general research (54 percent), travel inspiration (43 percent), and local recommendations (43 percent), with 84 percent of those users saying it improved their experience.

Media & Channel Strategy

Publicly documented channel-level effects diverge sharply by format and funnel stage. At the top of the funnel, AI-referred visits behave differently from conventional search traffic. Adobe’s data shows generative-AI referred visitors to U.S. retail sites spent more time engaged (8 percent higher engagement), viewed more pages per visit (12 percent more), and had a 23 percent lower bounce rate than non-AI traffic evidence that users arriving via an AI-generated answer tend to be more qualified or more deliberately directed than average search traffic. At the same time, Adobe found this traffic converted 9 percent less often than other traffic as of February 2025 though this gap had narrowed sharply from a 43 percent conversion gap recorded in July 2024, suggesting the channel’s purchase intent was strengthening as adoption matured. Adobe also found a pronounced device skew: 86 percent of generative-AI-referred traffic came from desktop devices between November 2024 and February 2025, compared with a 34 percent desktop share for e-commerce traffic overall indicating that, at this stage, AI-assisted shopping research was disproportionately a desktop behavior rather than a mobile one.

At the level of the open search results page itself, the Moloco-BCG index reports that when Google’s AI Overviews appear on a results page, 80 percent of those searches end without the user clicking through to any website a zero-click outcome that directly reduces the referral traffic available to brand and publisher sites regardless of ranking position. The same index reports that organic traffic to news publishers specifically had already fallen by approximately 26 percent, illustrating that the most exposed category under this shift is often third-party content publishers rather than brands themselves, even though those same publishers are frequently the sources AI engines cite when answering brand-related queries.

Business & Brand Outcomes

Because this case concerns an industry-wide structural shift rather than a single company’s campaign, there is no single, attributable P&L outcome to report, and no verified public information is available on the aggregate revenue or market-share impact of AI search on branded commerce as a whole. What is documented, from the sources above, are discrete, named outcomes: Adobe’s recorded year-over-year AI-referral traffic growth figures (1,200 percent, 1,300 percent, and 1,950 percent across the periods and events cited above) for its aggregate set of tracked U.S. retail sites; McKinsey’s $750 billion projection for U.S. revenue flowing through AI-powered search by 2028; and the category-level conversion findings reported by Adobe, in which electronics and jewelry recorded the strongest AI-referred conversion performance and apparel, home goods, and grocery recorded the weakest. No verified public information is available on which named individual brands have attributed a specific, disclosed revenue or market-share gain directly to AI search optimization efforts as of the sources reviewed here.

Strategic Implications

Several implications follow directly from the documented evidence rather than from inference. First, the unit of competition inside an AI-generated answer is smaller than it was on a search results page BCG’s observation that a large language model may surface only two or three options means that exclusion from an AI answer is a harder failure than a low search ranking, which still allowed a brand to be found with sufficient scrolling. Second, because McKinsey’s source-citation analysis finds that AI engines draw the majority of their answer content from third-party publishers, reviews, and user-generated content rather than brand-owned sites, a brand’s influence over its own AI-search representation is now exercised indirectly, through the broader information ecosystem that an LLM reads, rather than directly, through owned-channel content alone. Third, the low level of documented brand readiness only 16 percent of surveyed Fortune 500 consumer brand CMOs systematically tracking AI search performance, per McKinsey’s September 2025 survey suggests that, at the time of these reports, most large brands’ measurement infrastructure had not yet caught up to where consumer discovery behavior had already moved. Finally, the divergent industry exposure identified in the Moloco-BCG index — with travel, news, education, retail, and auto marketplaces more exposed than financial services or fintech — implies that the urgency and shape of a brand’s response should be calibrated to its specific sector’s documented exposure rather than applied as a uniform playbook across categories.

No verified public information is available on how these dynamics will resolve over a longer time horizon, including whether Gartner’s 2024 projection of a 25 percent decline in traditional search volume by 2026 will be borne out, given that several industry publications have since examined and contested that figure’s methodology.