The common phrase 'Google it' could soon be getting replaced by 'ask my agent', as more people begin to migrate over to an AI-only search-engine lifestyle.
With generic results being less fundamentally useful than personalized, contextual responses, traditional search engines are at risk of becoming a secondary research tool.
As more users default to AI platforms for retail experiences, what does the future of e-commerce look like?
In a CX Today interview with Jessica Keehn, CMO of SAP CX, she argues that 2025 would be the final year when most shoppers will default to 'Google it' when researching what to buy:
“This last year will be the last year where the majority of people started a search engine.”
Traditional VS AI-Powered Retail Experiences
With the scales tipping further in the direction of AI-assistants, this may lead shoppers to external AI platforms such as ChatGPT or Claude, or simply using the built-in AI assistants within retail websites.
Traditional Retail Search Experiences
These search engines rely on keyword matching, treating each customer query as a separate event and expecting the user to know exactly what to type to get its desired results.
As a result, incomplete, vague, or contextless searches may return with limited or irrelevant results to the customer.
And whilst it does heavily support valuable product discovery through structured input, results are typically ranked by popularity, global or national relevancy, or recency of when it was released, meaning results may end up untailored to a customer’s shopping needs.
AI-Powered Retail Search Experiences
AI-powered search engines work differently, using natural language processing to understand customer intent, context, and behavior without needing excessive and accurate context.
The system interprets what the shopper is trying to find from the information given and adjusts results accordingly, tailoring results based on previous history, location, behavior or preferences.
Functioning more like digital shopping assistants, they guide discovery, recommend alternatives, and help customers refine choices to reduce friction, improve relevance, and support faster decision-making.
“Product discovery is moving from search engines to AI recommendations," Keehn continues.
“When you're shopping with an agent, that agent actually knows you and your preferences, your result will be different than mine."
The Rise of AI Search Engines
The development of AI search engines in the retail infustry has only really gained momentum in the last few years; however, its expanded capabilities have been one driver in this increased popularity, no longer experimental but fully functional for everyday shopping.
One of the most prized advantages of AI search engines is their ability to drive personalized results quickly, a capability that traditional search platforms such as Google have historically led in.
In the early 2010s, Google had begun introducing personalization into the search engine, incorporating machine learning into ranking systems.
This led to the launch of RankBrain, one of Google’s first major AI-based search components, which helps users interpret unfamiliar queries and better understand intent.
Similarly around the same time, computer developers such as Apple had began introducing the early versions of voice assistants, such as Siri, increasing the popularity of conversational queries of keyword indexing.
This pushed forward natural language, and researching began to get faster.
As search engines continued to improve natural language understanding, interpreting context became more common that matching isolated keywords.
Towards the end of the decade, Google introduced the capability BERT into its search engine, significantly improving its ability to understand user context without needing excessive information or complete queries.
By 2022, generative AI search became more accessible to users after the release of LLMs, with OpenAI launching the earliest versions of ChatGPT, demonstrating basic conversational, context-aware responses instead of lists and links.
“Large language models have now sort of taken over the shopping experience,” highlights Keehn.
“Large language models are becoming a legitimate shopping channel.”
Despite its capabilities, traditional search engines continued to dominate the retail market, due to its familiarity and trust with customers over the past few decades.
However, to remain ahead of the game, search engines such as Bing began introducing AI into their search engines, combining both familiarity and automated summaries into search results.
In retail, many brands have begun introducing their own AI capabilities within their sites for customers to use, allowing them to research and discover new products without consulting traditional search engines first.
Now having had a four-year presence within modern-day research, AI search has evolved gradually from algorithmic ranking to machine learning-driven interpretation, then to generative, conversational systems.
Given the accessibility of AI today, it is plausible that users will see its eventual takeover of traditional search engines.
Will 2026 Be the Year That Shoppers Ditch Traditional Search Engines
SAP experts claim that 2025 could have been the final year that the majority of consumers used search engines as the default starting point for shopping research.
AI tools are being increasingly embedded within platforms to provide direct answers, summaries, and recommendations, with users being able to ask complex, natural questions with synthesized responses.
Even without AI, the past decade has seen an increase in younger users defaulting to social media apps such as Instagram to get product recommendations and reviews instead of search engines.
Furthermore, predictive systems have reduced the need for customers to search at all, as algorithms recommend content, products, and services based on previous searching behavior, with discovery happening passively through feeds and suggestions.
Keehn argues that there is a dramatic transformation in the consumer behavior, pointing out the decline in traditional search relevance.
