Perplexity's CEO Prediction vs. Reality: The Agentic Commerce Pivot

In 2024, Perplexity's CEO Aravind Srinivas made a clear prediction: ads would eventually be the company's core monetization engine. He believed this path could make Perplexity "really really profitable". That vision has now reversed. Last year, the company began phasing out ads, distancing itself from a model it once championed. This retreat is a direct pivot away from its earlier strategy and the CEO's stated goal.

The contradiction is stark against the broader industry. While Perplexity pulls back, OpenAI has embraced ads, recently testing them for free ChatGPT users. This sets up a major split in the AI sector's approach to profitability. Perplexity's executives cite user trust as the primary reason, arguing that ads could make people mistrustful of responses. The move also reflects a strategic shift toward a subscription model for high-value users, a path that requires less scale than mass-market advertising.

The tension highlights a fundamental question for AI companies: can they achieve profitability without the massive user bases that make advertising viable? Perplexity's user count, while growing, remains a fraction of giants like Google or OpenAI. By abandoning ads, the company is betting its future on enterprise sales and premium subscriptions, a strategy that may work for a niche but is a stark departure from the CEO's original, more aggressive monetization forecast.

The Practical Reality: Agentic Tools as Co-Pilots, Not Autopilots

The financial impact of agentic commerce is real, but it's incremental, not transformative. Walmart reported that its AI shopper assistant Sparky drives an average order value roughly 35% higher than for non-users. Amazon claims its Rufus generated $12 billion in incremental sales during its first season. These figures show agentic tools are effective at boosting revenue per transaction and capturing new sales, but they don't indicate a mass migration away from traditional shopping. The evidence points to a co-pilot role, not an autopilot. Analysis shows these tools are most effective for research and discovery, compressing the shopping funnel like an accordion. Consumers still pass through each stage of decision-making. As one writer notes, Rufus isn't acting as an autonomous agent but is really good search. It accelerates the process, but doesn't remove the need for human conviction in a purchase.

This sets up a clear economic reality. The winners in this space are the platforms with the deepest catalogs and trust, like Amazon and Walmart. Independent agentic commerce faces structural hurdles, including affiliate fee economics that break retailer profitability. The bottom line is that AI is changing how people shop, but the value capture remains firmly with the established e-commerce giants.

The Strategic Pivot: Browser Agents and the Data Battle

The battle between Amazon and Perplexity over browser agents is a proxy war for control over the user's shopping journey and the data it generates. Amazon sued Perplexity last November, claiming its AI agent was scraping the retail site without authorization by concealing its identity. The conflict escalated when a federal judge initially ruled in Amazon's favor, but Perplexity secured a temporary reprieve from an appeals court, allowing its agent to continue operating while the legal review proceeds. This clash is less about shopping convenience and more about who owns the data trail.

Perplexity's CEO has openly admitted the browser's true purpose: to collect comprehensive user data. On a recent podcast, Aravind Srinivas stated the browser is built to gather data on everything users do outside of its own app, including purchases and browsing. The goal is to build a more accurate user profile to better target ads. He believes this data will make ads more relevant, a trade-off he expects users to accept. This admission frames the browser not as a neutral tool, but as a data-gathering device for a future ad ecosystem.

Viewed through this lens, the strategic pivot is clear. The future isn't about replacing retailers with autonomous agents, but about platforms capturing the data from the entire shopping journey. Perplexity aims to be the central data hub, using its browser to profile users across the web. Amazon, meanwhile, is defending its walled garden to protect its own ad inventory and user data. The winner in this battle will be the platform that can aggregate the most complete picture of consumer behavior, turning it into a powerful asset for its own advertising and commerce networks.