As up to 70% of product discovery shifts toward AI-led interactions, brands, platforms, marketers are being pushed to rethink visibility, attribution, control in a rapidly evolving commerce ecosystem
The search bar once defined the architecture of e-commerce. It was the starting point of intent, the gateway to discovery and the anchor for performance marketing. Today, that centrality is being quietly eroded. In its place, a new layer is emerging where consumers no longer search for products but describe needs, delegate decisions and increasingly rely on systems that interpret and act on their behalf.
This shift marks a deeper structural change in how commerce functions. Discovery is no longer a linear journey driven by keywords and filters. It is becoming a dynamic process shaped by real-time signals, contextual understanding and continuous interaction between consumer and system. The interface is changing, but more importantly, so is the logic that underpins it.
What makes this transformation notable is not just its scale but its speed. The convergence of large language models, unified data layers and conversational interfaces is accelerating a transition that was once considered gradual. The result is an ecosystem where intent is captured differently, decisions are influenced differently and increasingly, transactions are executed differently.
From search queries to interpreted intent
According to Abhijat Shukla, VP Data Science at WebEngage, around 40 to 45 percent of Indian consumers are now using AI tools during their purchasing journey. WebEngage is a customer data and engagement platform that enables personalised marketing at scale to improve customer lifetime value and ROI. More tellingly, 95 percent report that AI influences their decision-making in some capacity. Close to 70 percent of shoppers say they plan to use AI in future shopping decisions, pointing to a clear directional change in behaviour.
“When looking at the overarching trends, it’s anticipated that AI will impact approximately 40 to 50 percent of purchase journeys,” Shukla says. “In the coming year or two, we will see this number increase significantly around 60 to 70 percent in the next two to three years. Product discovery in India is transitioning from being search-based towards becoming increasingly AI curation-based, where consumer intent is interpreted in real time and acted on accordingly.”
This is not a marginal shift. It is a redefinition of how intent itself is captured. Instead of typing structured queries, consumers are increasingly expressing needs in natural language or relying on systems that predict those needs based on behaviour and context.
Globally, similar patterns are emerging. Pankaj Srivastava, Founder and CEO of UnoSearch, a ROI-focused PPC, SEO, and GEO company, estimates that 38 to 42 percent of product discovery today is already AI-influenced. He points to a sharp rise in generative AI usage in commerce, with 64 percent of consumers having used such tools for shopping, up from 51 percent just a year ago.
“Shoppers arriving from generative AI sources show 10 percent higher engagement, 32 percent longer site visits and a 27 percent lower bounce rate,” he says. “For a
growing cohort of digital-native shoppers, AI isn’t a layer on top of discovery. It is the discovery process.” He expects this share to cross 65 to 70 percent in the next 24 to 36 months, driven by the rise of AI-native interfaces, shifting consumer behaviour and growing inefficiencies in traditional search.
The failure rate of keyword-based discovery is becoming a catalyst for change. As Srivastava points out, 86 percent of consumers frequently reformulate their queries
because they cannot find relevant results, while 66 percent of frustrated shoppers move to alternative platforms. In that context, AI-led discovery is not just an enhancement. It is a correction.
The quiet rebuild of commerce infrastructure
While the shift is visible at the interface level, its foundations lie deeper in the architecture of commerce platforms. What appears to be a simple move from search bars to chat interfaces is in reality a full-stack transformation.
Shukla notes that marketplaces and retail platforms are increasingly adopting unified real-time data layers that combine behavioural, transactional and contextual signals. These systems allow AI agents to interpret intent more accurately and deliver personalised recommendations without requiring explicit input from users. Proactive discovery is becoming a defining feature. Instead of waiting for a query, platforms are beginning to surface products based on inferred intent, past behaviour and contextual cues. This reduces friction and shortens the path to purchase.
Srivastava describes this transformation as a layered rebuild. At the core is the shift from keyword-based indices to semantic retrieval systems that understand meaning rather than just matching words. This enables platforms to respond to complex, natural language inputs that reflect real-world intent.
The catalogue layer is also being restructured. Traditional product listings designed for human browsing are being replaced with structured, attribute-rich data that can
be interpreted by AI systems. This includes contextual attributes such as use cases, compatibility and situational relevance. The personalisation layer is evolving from
static profiles to real-time context synthesis. Systems now combine multiple data points including browsing behaviour, purchase history, time of day and even environmental factors to generate recommendations in real time.
According to Srivastava, AI-assisted purchase decisions are happening 47 percent faster, while returning users interacting with AI systems are spending 25 percent more. The interface itself is undergoing a visible transformation. Conversational systems are replacing navigation-heavy journeys, allowing users to express intent in
natural language rather than navigating categories and filters. However, Srivastava cautions that simply adding a chatbot does not constitute transformation.
“Too many brands have bolted a chatbot onto their existing site and called it AI-first. That’s like putting a GPS screen on a horse cart,” he says. “True agent-led discovery means the interface curates choices and manages the journey without requiring the user to navigate.”
Commerce without browsing
The impact of these changes is already visible in live deployments. Fynd, an AI-native retail technology company has processed over 4.3 million customer interactions using AI models and currently handles more than 750,000 interactions per month. Its systems have served over 1.1 million users across multiple channels including web, mobile apps and messaging platforms.
Farooq Adam, Founder of Fynd, sees this as a fundamental shift in how commerce is experienced. “Consumers today are no longer just navigating platforms. They are interacting with them in more intuitive, real-time ways. With the diversity and scale of Indian retail, the ability to understand intent and respond instantly within a unified system is becoming the defining layer of commerce,” he says.
In fashion, AJIO has deployed conversational discovery where customers describe intent through open-ended prompts such as outfit requirements. Nearly 90 percent of interactions are driven by discovery, with a 99 percent positive response rate. This suggests that consumers are not just adapting to conversational interfaces but actively preferring them. During peak sale periods, Being Human Clothing handled over 115,000 interactions in three days using AI systems while managing a 2.5 times surge in traffic with full uptime. Post-sale engagement also remained elevated, with daily message volumes increasing 2.8 times.
In healthcare, Netmeds has managed over 850,000 customer queries through conversational interfaces, achieving 96 percent favourable sentiment. These examples
indicate that AI-led interactions are not limited to discovery but extend across the entire customer journey.
Sreeraman Mohan Girija, Founder of Fynd, reinforces this view. “Consumers already expect the kind of personalized instant interaction that only AI-native infrastructure can deliver. The results we are seeing across 1.1 million users tell us conversational commerce isn’t a feature. It’s the next primary interface for retail,” he says.
Retail media at an inflection point
As discovery shifts away from search, the implications for retail media are becoming increasingly difficult to ignore. The current model, built on keyword bidding,
sponsored listings and display placements, is designed for environments where users actively search and browse. In an agent-led ecosystem, those assumptions no longer hold.
Shukla believes that advertising will need to evolve to remain relevant in this new context. “As the agents facilitate the entire discovery process, the advertising models
used today will not work effectively in an agent-led environment,” he says. “In this new landscape, advertising will pivot towards sponsored recommendations that are
woven into AI-driven conversations. Rather than relying on banners or paid listings, brands will emerge as part of relevant, intent-focused suggestions.”
This represents a shift from visibility to relevance. Instead of competing for placement on a results page, brands will need to earn inclusion in recommendation sets generated by AI systems.
Srivastava argues that the implications are even more profound. “The current retail media model is essentially an auction for human eyeballs at the moment of intent,” he says. “Now remove the human from that loop. An AI agent doesn’t scan a page. It doesn’t click a sponsored result. It synthesizes an answer.”
In such a scenario, traditional ad formats lose their visibility. Srivastava suggests that new models will emerge where brands pay to be included in the agent’s consideration set based on intent signals rather than keywords. Attribution models will also need to evolve, moving away from clicks and impressions toward outcomes such as recommendations and agent-initiated transactions. He also highlights the growing importance of structured data. In an agent-mediated ecosystem, the quality and completeness of product data become critical determinants of visibility. Brands that invest in making their data machine-readable are likely to gain an advantage.
Transparency is expected to play a key role as well. As AI systems begin to incorporate sponsored elements into recommendations, disclosure and regulatory frameworks will become increasingly important in maintaining consumer trust.
The rise of autonomous commerce
The logical extension of AI-led discovery is autonomous commerce, where systems not only recommend products but also initiate and complete transactions. While this
may still be in its early stages, industry estimates suggest that it is closer than it appears.
Shukla estimates that 20 to 30 percent of e-commerce transactions could be initiated and completed by AI agents over the next five years without direct user input. This
shift is likely to begin with routine purchases such as groceries and household essentials, where decision complexity is low and preferences are well established.
Srivastava projects a higher share, suggesting that 45 to 52 percent of transactions could be agent-driven by 2030. He breaks this down into categories, noting that
replenishment commerce is already moving toward automation, while mid- consideration purchases are likely to follow over the next few years.
Higher-consideration purchases may take longer to transition, as they involve greater emotional and financial stakes. However, the direction of travel remains clear. The implications of this shift extend beyond convenience. As AI agents take on a larger role in decision-making, the dynamics of trust, control and accountability will become increasingly important. Consumers may delegate routine decisions, but they are likely to remain involved in more complex ones.
The transformation underway in e-commerce is not just about replacing search with conversation. It is about redefining how discovery happens, how decisions are made and how value is created across the ecosystem. For platforms, the challenge lies in rebuilding infrastructure to support intent-led interactions. For brands, it involves rethinking how visibility is earned in an environment where algorithms mediate discovery. For marketers, it requires new models of attribution and engagement that go beyond clicks and impressions.
The search bar may not disappear overnight, but its dominance is already being challenged. In its place, a more adaptive and less visible layer is taking shape, one that
is likely to define the next phase of digital commerce.
Click here to read more: e4m
