Unilever Just Paid $1.2B for Grüns. We Had the AI Diagnostic Data the Same Day.

2026-04-11 · Source: LLM on Medium · Field: Business & Management — Marketing, Branding & Advertising, E-commerce & Digital Commerce, Corporate Strategy & Leadership · Depth: Intermediate, medium

Summary

Unilever acquired the supplement brand Grüns for $1.2 billion, but an AI diagnostic run on the day of acquisition revealed a CODA score of 8/100, indicating the brand was eliminated from buying recommendations on both ChatGPT and Perplexity. This low score resulted from a four-turn decision-path probe, where Grüns failed to maintain position against competitors and did not meet specific criteria like clinical evidence, leading to its exclusion from final recommendations. The analysis suggests Grüns has an awareness problem, appearing in early AI conversations, but a critical positioning problem, as its current "greens in gummy form" messaging lacks the clinical evidence AI models prioritize for comparative judgments in the supplement category. This highlights a broader issue where brands optimize for visibility but lose at the decision stage in AI-influenced commerce.

Key takeaway

For AI Product Managers evaluating brand acquisition targets or optimizing existing portfolios, your strategy must account for AI recommendation performance. A high CODA score, driven by structured clinical evidence and robust entity infrastructure, is crucial for winning AI-assisted purchase decisions. Neglecting this "decision-layer work" means losing significant sales volume in AI-influenced categories, even for brands with strong traditional awareness and distribution.

Key insights

AI-driven purchase decisions prioritize clinical evidence over lifestyle positioning, impacting brand recommendations.

Principles

• AI recommendation requires decision-layer optimization, not just awareness.
• Clinical evidence is a Type 1 filter for supplement category AI.
• AI referral analytics often miss lost decision-stage buyers.

Method

A four-turn decision-path probe, mirroring a buyer's journey from research to purchase, is run across AI platforms to assess brand recommendation performance and generate a CODA score.

In practice

• Structure ingredient-level clinical evidence for AI extraction.
• Build entity infrastructure (Wikipedia, Wikidata) for AI knowledge bases.
• Create decision-instruction content addressing AI criteria language.

Topics

• AI Recommendation Performance
• CODA Score
• Decision-Path Probe
• Clinical Evidence
• Supplement Category

Best for: AI Product Manager, Product Manager, Entrepreneur, Marketing Professional, Consultant, Executive

Related on AIssential

• This Week in AI: The Next-Gen Recommendation Experience — Recommendation systems are underutilized revenue drivers, evolving into complex sequence prediction and foundation model applications.
• AEO Changes What Content Must Do: Help Buyers Decide — AI-powered search necessitates decision-driving content that explicitly frames comparisons and trade-offs, not just explanations.
• ChatGPT Was 76% of the AI Market. It Just Fell to 54%, and Quality Isn’t Why. — In consumer AI, distribution ownership, not just model quality, is the primary driver of market share.
• 🥊 SaaSpocalypse hits world capital market. — AI models are rapidly advancing in cybersecurity, financial automation, and agent-based platforms, while disrupting traditional software business models.
• Salesforce Ventures, Echo Health Ventures back Optura's $17.5M series A to track AI performance - Fierce Healthcare — Quantifying AI's business value is critical, especially in healthcare, where adoption outpaces measurable ROI.
• Your Brand Reputation Precedes You With AI, Whether You Like It or Not — AI systems exhibit "Narrative Gravity," perpetuating established brand stories and amplifying existing digital authority.

Open in AIssential →

Editorial summary, takeaway, and curation by AIssential. Original article published by LLM on Medium.