Spangle AI Says Retailers Must Adapt as Personal AI Assistants Reshape Product Discovery

Personal AI assistants are changing where shopping decisions happen, and retailers need intelligent agents of their own to keep up, according to Maju Kuruvilla, founder and CEO of Spangle AI and a former Amazon technology leader.

For 25 years, retailers built their digital businesses around drawing customers to their websites and helping them find products once there. Search, advertising, navigation, recommendations and personalization were all optimized around that model. Now, consumers can describe what they're trying to accomplish and let an AI assistant evaluate alternatives and recommend products. Kuruvilla argues that customer needs haven't changed-people still want the right product, at the right price, from a brand they trust-but how they express those needs, and who helps them decide, is shifting.

The challenge for retailers goes beyond visibility. Many have accumulated hundreds of recommendations from analytics, marketing and merchandising systems without the capacity to act on them, as teams work through separate systems and workflows. Kuruvilla contends that more AI traffic or citations don't necessarily mean more revenue, and that AI investments should be judged on commercial outcomes rather than visibility metrics.

Spangle aims to address this by combining a retailer's product, inventory, brand and merchandising context with AI agents that help identify opportunities and execute approved actions, while the retailer stays in control. Express, which had gathered more than 800 recommendations, used Spangle's AI Discovery product to translate opportunities into shopping experiences. In the first full month after launch, Express reported a 14.5-times increase in AI-attributed revenue compared with the preceding 30 days, along with 6.9-times growth in AI-referred sessions and a 6.7-times increase in conversion rate for those sessions-before-and-after results rather than a controlled measure of incremental lift.

Kuruvilla says retailers should start with a meaningful business problem, establish a baseline and test whether the technology improves the outcome before expanding. "Our goal isn't to add another AI tool to the retailer's technology stack," he said. "It's to help retailers make better decisions and turn those decisions into better business results."