Major retailers are taking opposing stances on AI agents, autonomous systems that search, compare, and purchase on behalf of customers.

Target and Walmart are embracing the shift, partnering with OpenAI to enable shopping directly within ChatGPT. Amazon is playing defense, blocking Perplexity AI's agentic shopping from its platform to protect $17.7 billion in quarterly ad revenue at risk if agents replace direct marketplace visits.

Both strategies make sense given their market positions. Target's and Walmart's partnerships aims to preserve customer relationships by shaping how agents interact with its platform. Amazon leverages dominance to wall off its traffic. But neither approach works for most retailers. Few have Walmart's scale to negotiate favorable terms or Amazon's deep pockets to fight legal battles. Even Amazon may find resistance futile, with Google launching AI-powered shopping and leveraging its dominant position, the choice may soon be: allow agent access or lose traffic to competitors who do.

Most retailers need a third approach: building foundational capabilities that create value regardless of how the agent landscape evolves.

Navigating the Unknown

AI agent technology remains early-stage, struggling with context, complex preferences, and multi-step transactions. Yet capabilities are advancing rapidly. The question isn't whether third-party agents will access retail platforms, but when, and under what terms.

When agents capture customer relationships, retailers risk being forced to compete purely on price, losing brand equity, customer insights, and retail media opportunities as algorithms shield customers from the suppliers of goods.

Data as Strategy

The strategic response is clear: invest in data quality that delivers value today while positioning for an AI-enabled future.

AI agents prioritize structured, machine-readable data. While AI systems can technically interpret marketing text or product images, doing so is expensive and error-prone. Agents optimizing for speed, accuracy, and cost will favor retailers offering clean, structured information over those requiring resource-intensive interpretation.

When agents compare thousands of products instantly, they will have zero tolerance for missing fields or ambiguity. If critical attributes are incomplete, buried in marketing copy, or only visible in images, agents will skip your products for competitors with explicit data.

To remain competitive, retailers must ensure high quality product data but also consider structured fields detailing:

• Functional tags: induction-compatible, vegan leather, IP68 water-resistant;
• Standardized use-cases: suitable for sensitive skin, intended for indoor use;
• Safety & compliance information: age-suitability, dietary/allergen info, sustainability certifications, material / component / ingredient lists.

Improving the breadth, depth, and quality of product data enhances search, recommendations, and inventory management today while also providing the foundation for advanced internal AI capabilities tomorrow. Crucially, superior data provides retailers with options. This allows them to engage with multiple AI agents simultaneously rather than betting on a single platform. This prevents lock-in, enables strategic experimentation, and preserves control of customer relationships.

This isn't theoretical. Walmart invested in AI-driven data quality improvements across 850 million product attributes, which simultaneously optimized their existing operations and enabled their OpenAI partnership, exactly the kind of optionality that better data provides.

The retailers who thrive won't be those who try to predicted the future, but those who built foundational capabilities that create value today while remaining ready for however the agentic future unfolds tomorrow. In an era of uncertainty, adaptability is strategy.