Jennifer Peters, Director of DTC, MarTech, and Digital Compliance at OLLY, a supplement brand owned by Unilever, has quietly built one of the most practical agentic commerce playbooks in the industry (Retailgentic). Peters, a self-taught Shopify developer, has been leveraging Claude and model context protocol (MCP) to optimize OLLY's digital presence for AI agents. Key wins include catalog feed enhancement, adding FAQ sections to product detail pages by repackaging existing content, and working with partners to push back AI-ready attributes that are not customer-facing (Retailgentic).
OLLY's approach reflects a broader shift in e-commerce strategy: the website has evolved from representing 7 or 8% of the business to becoming the brand authority source that AI agents consult first—whether for OLLY's own site, Walmart's Sparky, or Amazon's Rufus (Retailgentic). Peters emphasizes that agentic e-commerce optimization (AEO) is fundamentally good SEO: clean schemas, structured data, and organized marketing and DAM stacks matter more than bolting AI on top of broken foundations. Her philosophy—"Don't do AI to do AI"—means keeping the same business objectives and using AI to reach them faster, rather than chasing novelty (Retailgentic).
For commerce practitioners watching the agentic shift, Peters' work signals that competitive advantage comes from unglamorous groundwork: connecting data systems, verifying recommendations across tools like Profound and Contentsquare, and treating agent-readiness as a natural extension of existing customer-facing optimization. The counterculture risk she flags—growing backlash against data centers and AI-heavy creative—suggests that brands balancing innovation with restraint may win customer trust in the coming wave (Retailgentic).