The end of online search as we know it?

For decades, online visibility has followed a familiar logic: users type keywords, search engines return blue links and businesses learn how to optimise their way to the top. Today, that model is breaking down. Conversational artificial intelligence is rapidly rewriting how people search for information, how search technology surfaces results and how value flows within the digital economy.

Earlier this year, the MIT Technology Review declared that AI will mean the end of internet search as we’ve known it. As 2025 has progressed, we’ve watched this prediction hold up.

“If you’re a business trying to stay visible online, if you are a platform enabling discovery, if you are a policymaker thinking about competition, this is reshaping your business reality right now”– Laure Claire Reillier, co-founder and COO of Launchworks & Co

From keywords to conversations

With around 700 million weekly users according to OpenAI, ChatGPT is already a digital phenomenon. Since ChatGPT’s launch, competition has been heating up with conversational tools like Perplexity, Gemini, Grok and others joining the fray. The habits that users form in these AI-driven environments fundamentally reshape their expectations of other digital experiences – for example, search. Younger audiences in particular are increasingly bypassing traditional search engines like Google altogether, discovering products and services through social feeds or conversational interfaces. 

“Nobody uses keywords. We use long strings and conversations,” says Jon Earnshaw, co-founder and Chief Product Evangelist at Pi Datametrics, adding, “the old days of optimising a webpage for a keyword to get onto page one of Google [are] gone forever.” Short, transactional queries are no longer the norm. 

Instead, users are increasingly engaging in long, multi-turn conversations with AI. People are no longer searching for information or items simply by name or category, such as ‘best men’s blazer’; they are asking detailed, contextual questions that reflect intent, budget and personal preferences. Jon illustrates the contrast: “‘What are the best men’s blazers for winter 2025 that are under 300 quid and suitable for smart casual.’ That is a prompt.” Search technology, he argues, has shifted from ranking pages to synthesising answers based on information-rich content.

The fundamental differences of AI search

The mechanics of AI search combine old and new elements in unfamiliar ways, which accommodates these behavioural changes and creates an entirely new user experience. Professor Alex Yang, professor and area chair of management science and operations at the London Business School, describes AI search as a complex product: “The first part is information retrieval. The second part is information synthesis or aggregation.”

Unlike classic search engines, which surface a range of links, AI systems compress information into a single, narrative response. “We are seeing a lot of changes on how the value chain is restructuring from both the demand side, as in how people are doing search differently, and from the supply side, as in how the information is presented and aggregated,” Alex explains. “From the supply perspective, traditionally, we had a bunch of links; now we have a summary.”

“The current AI search [paradigm] may actually be a transient phase because we’re moving so quickly with AI” – Alex Yang, Professor & Area Chair of Management Science and Operations, London Business School

How prompts are interpreted, how sources are selected and how summaries are generated all remain opaque. Initially, Perplexity performed traditional search and retrieval using Microsoft’s Bing engine, before presenting the top results in its conversational interface, Alex recounts. (Perplexity.ai says that they have since changed this approach.) Other AI systems, such as ChatGPT, don’t use preexisting search technology, but rely on limited, curated sources. For Alex and others, this raises questions about whose content is visible and whose is excluded.

“The current AI search [paradigm] may actually be a transient phase because we’re moving so quickly with AI,” Alex further notes.

A widening gap for small businesses

For small and medium-sized enterprises (SMEs), the shift to conversational AI-based search is both invisible and massively consequential. Polly Dhaliwal, chief operating officer of Enterprise Nation, believes that most SMEs are not thinking about AI search at all. “Not everyone is in the digital space,” she reminds us.

For many founders, immediate concerns dominate: cash flow, hiring, access to markets. Search strategy rarely makes the list. Yet Polly warns that this lack of awareness risks deepening the existing “digital skills gap”, particularly for regional economies and high-street businesses – many of which don’t have their own website or even a social media page. She also notes that female founders, who already face major hurdles in accessing capital funding, could be at a further disadvantage on an unlevel playing field. “Government and organisations need to do a lot more on simplifying the message [about AI and] making it more inclusive for small businesses,” she says. “Not enough businesses know about the issues that are about to hit them.”

“Not enough businesses know about the issues that are about to hit them.” – Polly Dhaliwal, COO, Enterprise Nation

Enterprise Nation’s work with small businesses across the UK keeps them in close contact with the reality on the ground. Polly sees an urgent need to help SMEs from falling even further behind: “Small businesses make up 98% of the UK economy. So I feel we need to do something – quick.”

The broken bargain of online content

This rapid change is creating new economic tensions online as well as offline. For years, content creators accepted being crawled and indexed because search engines sent traffic in return. AI search disrupts that exchange.

According to Alex: “Studies show that once Google provided an AI summary, the click-through rate has dropped dramatically.” For information-led businesses, this is existential. “All publishers are in trouble,” thinks Jon. He notes that many publishers have been exploring a range of survival strategies, from advertising to affiliate models, and are now experimenting with  licensing agreements, and new compensation mechanisms sometimes at the same time as litigations! 

For his part, Alex believes that the landscape will continue to evolve, noting that, “in the long run, Google will realise this is a problem.” He also suggests that two-way payment models – where content used in AI summaries is compensated, and click-through traffic still creates value – could help realign incentives.  “If I allow Google to display my content in the AI summary of organic search results, then I should be paid a very small fee. But if a customer clicks through the AI summary to my website, maybe I will pay Google,” he says by way of example. For now, however, the economic framework remains unresolved.

Optimising for intent, trust and originality

The economics may be uncertain, but businesses can already consider how to adapt their content strategies in response to AI search. “Today, if you want to get Gen Z onto your site, you’ve got to write fast content,” Jon thinks. “It’s got to be seen to be firsthand, authentic, social and trusted.” He sketches what this might mean in practice with an example: purchasing a new barbecue. “Before, we just optimised for ‘best gas barbecue’ or ‘best charcoal barbecue’ or ‘best barbecues’,” he says. “[Now] you have to anticipate not just the first question, but the second and the third follow-up questions.” 

Jon also recommends connecting information together so that AI can retrieve it easily, which doesn’t necessitate “stuffing it onto one page” as per traditional SEO best practices. Instead, he suggests focussing on trust. “If you are genuinely authentic and firsthand and seem as an authority, then you will get the citations [in conversational search]. It’s about getting your brand mentioned in the conversation,” he thinks.

In this new paradigm, depth, originality and authority may indeed matter more than volume. As Jon says, AI systems reward content that reflects genuine expertise, firsthand experience and verifiable authorship. According to Polly, many SMEs have been enthusiastic adopters of AI tools for generating marketing content, but are finding, perhaps paradoxically, that it underperforms in the age of AI search. In her opinion, “that’s because it’s not trustworthy, it’s not transparent and it looks like it’s been generated by AI.”

For his part, Alex also reinforces the potential trade-offs of AI-generated content: “Generative AI has a super big impact on content generation, [and] we have to be very careful with that. It lowers the cost of content generation, but what kind of content does it generate?” His recent research into this question has yielded interesting results. One study used data from spring 2023, during the month when OpenAI disabled ChatGPT in Italy after an order from the country’s government data protection agency. The study examined social media content among comparable small businesses in different regions. “During that period, Italian Instagram posts were more diversified,” Alex reports. “[The posts] actually collect more likes when they’re more unique [and] more original.”

Navigating visibility on an AI-first internet

The shift to AI-mediated search is not a future scenario. It’s already reshaping visibility, competition and value creation.

Polly’s message to SMEs is pressing but encouraging. Most businesses don’t have the resources to rewrite vast archives of keyword-optimised content; instead, she suggests narrowing in on a selection of top-performing pieces and adapting them for AI search. Beyond that, she recommends a return to brand basics. “If you’re a founder, I really urge you [to] go back to your vision, go back to your mission,” she says. “That ‘why’ creates the content. What is the problem you’re trying to solve for end users? That is the content that will then lead them to your product, your service.”

“We’re talking about not optimising content for keywords, but optimising for intent” – Jon Earnshaw, Chief Product Evangelist, Pi Datametrics

For businesses large and small, the challenge is to move beyond tactical SEO and towards strategic clarity – again, investing in trust, originality and human expertise. According to Jon, businesses need to start thinking about “not optimising content for keywords, but optimising for intent.” 

Alex thinks that some firms may need to make larger strategic pivots for the next era, depending on their product or service. “If you offer something that is easily substitutable by generative search, then you are in trouble. If you manage to offer something complementary, this could actually benefit you,” he says, with yet another reminder of the uncertainty and opportunities that lie ahead.

Shape the debate

This panel with Jon Earnshaw, Polly Dhaliwal, Laure Claire Reillier and Alex Yang took place at the Platform Leaders event organised by Launchworks & Co, held on 18 November 2025. To continue the conversation, the Platform Leaders website has the full agenda, session recordings and articles. Join the Platform Leaders community today to keep up with future events and further insights. Explore the agenda, watch session recordings and join the community to stay up to date with future events and insights. You can watch the full event by playing the video below.