AI is reshaping online search in two distinct but overlapping ways. Both reduce friction for consumers, but they increase friction for businesses.
The first is that large language models (LLMs), such as ChatGPT and Microsoft Copilot, are starting to displace search engines as where consumers go for answers. People type in a search query and receive a synthesized response from vast amounts of text. Sources and brands may be cited selectively or not at all. The second is Gemini-powered Google Overviews are diluting website traffic. These overviews sit on top of traditional search and provide an AI-generated summary before the familiar list of links.
These developments represent a significant change to how businesses are found and accessed. They have fundamental impacts for business strategy and for how organizations operate.
In this article we highlight three of these shifts and suggest how companies can navigate this modern discovery landscape.
Shift 1: AI Recommendations are Becoming More Influential
Companies spend billions on advertising to drive brand recognition. Procter & Gamble (P&G), for example, are among the highest spenders in the U.S. pushing the recognition of consumer brands like Pampers, Tide, and Gillette and spending around $9 billion annually. Yet decades of consumer research show that while people tolerate advertising, they trust recommendations. And today, that trust is shifting away from human sources (friends, family, salespeople, influencers) and toward algorithmic ones (primarily, LLMs).
That suggests that consumers are becoming less influenced by the push of advertising and more influenced by the pull of AI recommendations. Instead of relying on the brand-driven messaging of large advertisers, consumers are leaning on the data-driven messaging of AI with questions like: “What’s the most effective and least expensive brand of diapers?” Or: “Which washing liquid would you recommend for a restaurant?”
As a result, many companies are reassessing how and where they allocate their advertising dollars in response to this changing consumer behavior.
Take for example, a major European online retailer we’ll call Nordpay. It has positioned itself directly in the “AI recommendation” flow by rolling out a personalized AI-driven stream of products that it shows each customer. It also uses AI systems to create content, recommendations, and interactions in real time to improve the shopping experience and to personalize marketing. Instead of increasing its advertising budget to “keep up” with the competition, Nordpay is reallocating its spending away from external agencies and toward AI-enabled in-house production.
Specifically, as one advertising executive explained to us: “We’ve reduced our advertising spend by 11% while still producing more marketing output. We’ve cut agency spend by approximately 25%, shifting work in-house with gen AI. And we’ve replaced much of our image production workflows with tools like Midjourney, DALL-E, and Adobe Firefly, shrinking the image development cycle from approximately six weeks to around seven days.”
What Companies Can Do:
Review your advertising budget to see if it’s being wisely spent. Ensure you use consistent and specific language across all content, so LLM-powered, LLM-mediated, and LLM-based tools associate these specific ideas with your brand. Even better, create concepts with your name attached (e.g., “The Acme Index,” “The Smith Method”) so AI systems associate your ideas with your brand. Reallocate your advertising budget toward AI-native recommendation channels and build in-house generative AI capabilities to accelerate and lower the cost of creative production. Shift from campaign-based advertising to continuous, always-on experimentation.
Shift 2: SEO and Website Design Matter Less and Less
Traditional search followed a clear pattern. A customer entered a query; search engines returned a ranked list of websites; and the customer clicked on a few links to find the best answer to their query. This process created a multi-step customer conversion pathway in which businesses had room to differentiate through website design, useful pages, testimonials, and examples.
This is now collapsing. When customers ask an AI tool or default to just looking at Google AI Overviews for product and service recommendations or professional insights, they receive a fully formed answer—not a set of links to explore.
One recent study found that when AI summaries appear in search results, users clicked on ranked websites only 8% of the time, compared to 15% without AI. This is a 47% reduction in clicks. For some publishers, click-through rates have alarmingly dropped by as much as 89% as the customers’ exploratory stage disintegrates and branded touchpoints, where organizations once shaped competitive advantage, largely disappear.
Consider the experience of one large private U.S. health insurance provider we’ll call HSure. It has found that the traditional brand recognition pathway for consumers is being compressed by LLMs. Search engines once directed potential consumers to HSure’s educational pages, policy explainers, and comparison tools, often across multiple website visits before a purchase decision was made. These repeated interactions established trust, conveyed regulatory nuance, and positioned the firm as a credible expert in a complex domain.
Increasingly, consumers ask just direct their questions LLMs such as ChatGPT. The model generates a comprehensive response synthesized from HSure’s policy descriptions, benefit structures, and regulatory explanations, alongside similar material from competitors. It arrives without a direct link to HSure’s website.
As HSure’s Head of Operations, Julia, explains: “Our internal analysis has shown that information that previously required 15 to 20 website visits across the customer research journey is now delivered in a single LLM-generated response. Our brand recognition is removed from the customer relationship, and we lose not only traffic and conversion opportunities, but our role in guiding high-stakes decisions about health, risk, and financial protection.”
What Companies Can Do:
Shift your content plans from optimizing pages to engineering recall. Share original, organization-generated data, first-hand experience, strong points of view and named in-house experts with credentials. LLMs are far more likely to recall: “According to HSure’s Healthy Plus Survey…” than “Our study suggests…” Attach real experts’ names, credentials, and biographies to content. Create signature concepts, benchmarks, or brand-named frameworks that become shorthand for your thinking. Structure content in clear, quotable language so AI models can easily reproduce it. Measure success not only by traffic, but also by whether your brand and experts are mentioned, paraphrased, and associated with key ideas inside AI-generated responses.
Shift 3: Marketing has a New Audience
When customers ask, “What’s the best accounting solution for my small business?” Claude delivers synthesized recommendations in a conversational format, often eliminating the need to click through to links or brand pages. The result is that AI, rather than the customer, controls the first impression and traditional marketing loses its punch.
A major product review and affiliate website, here called Product Insight, has witnessed this. Marketing had invested heavily in featuring comprehensive product comparisons and expert reviews for categories ranging from laptops to kitchen appliances to drive revenue.
However, because consumers can now ask Gemini-enabled Google Overviews questions like “Which vacuum cleaner is best for pet hair?” Product Insight’s traffic data reveals that their historically highest-value pages have seen traffic decline by 67%. Google Overviews now appear for 78% of their core product queries.
This demonstrates that when Google introduced AI Overviews, it did more than alter search results—it changed the role of marketing. Increasingly, brands must persuade not only customers but also the algorithms that mediate their interactions.
Consider the U.S. publishing company we’ll call Henry Smith, which owns a home and lifestyle platform delivering practical guidance on décor, gardening, DIY projects, cleaning, and entertaining. Marketing emphasized everyday advice to help readers create living spaces and it combined expert-backed tutorials with trend insights and curated product suggestions. Its business model depended on ranking highly for queries like “what’s the best noise-cancelling headphones under $200,” converting intent into affiliate revenue.
AI systems now synthesize those rankings directly in search results, often without a click, link, or brand message. The algorithm—not the publisher—frames the recommendation.
In response, Henry Smith has redesigned marketing itself. As its Chief Growth Officer, Michael, explains: “We now structure content so machines can parse authority and expertise. We’ve invested in schema (a standardized vocabulary that labels content in a machine-readable way), increased authorship signals (indicating who created the content and why they are qualified to do so), and provided clean data architecture (how content is structured, organized, and coded so that algorithms can easily interpret it). And we’ve accelerated efforts to build direct audience relationships through newsletters and branded search.”
For Michael the strategic lesson is stark: “Marketing is no longer solely about influencing human perception. In an AI-mediated marketplace, the first customer is the algorithm.”
What Companies Can Do:
Write content that clearly defines concepts and uses structured explanations such as steps, and definitions. State conclusions plainly (LLMs love clarity) and make your brand easy to cite with a clear brand name, clear positioning (“X is a Y that does Z”) and consistent description everywhere (site, LinkedIn, media, Wikipedia-style profiles). Ensure your brand appears consistently in industry publications, expert interviews, conference agendas and author bylines. Encourage third-party discussion about your brand (reviews, comparisons, case studies). LLMs infer importance from frequency plus consistency across sources.
. . .
For more than 20 years, companies have relied on search engines as the backbone of their competitive advantage. They honed SEO plans, built expansive content libraries, and invested heavily in capturing clicks, attracting website traffic, and converting interest into revenue.
That era is ending with the advent of artificial intelligence. Consumers no longer need to visit websites—and brands are becoming invisible to consumers. As answers replace links and synthesis replaces exploration, visibility is no longer earned through clicks but through a branded presence inside AI systems.
The winners in this new AI-dominated landscape will be those who treat AI not as a channel to optimize, but as an audience to influence. That requires clarity of expertise, consistency of signal, and sustained brand building beyond search.
