Ulta Beauty Teams Up With Google Gemini on Agentic AI

Ulta Beauty is partnering with Google Gemini to roll out agentic commerce in Google search AI mode and the Gemini app within the next month, enabling recommendations, comparisons and streamlined checkout for eligible purchases. The company is also launching Ulta AI on Gemini Enterprise, using insights from its 46 million members to personalize shopping. The move strengthens Ulta’s digital and AI capabilities, but the near-term market impact is likely limited.

Analysis

ULTA is the clearest near-term beneficiary because this moves the company one step closer to owning the highest-intent moment in beauty: product discovery to checkout. The second-order effect is not just incremental conversion; it is lower customer acquisition friction and better monetization of its loyalty graph, which should improve LTV/CAC at the margin over the next 2-4 quarters. If the integration meaningfully shifts browsing into conversational commerce, ULTA can reduce dependence on paid search and partially insulate itself from higher auction costs.

GOOGL benefits less from revenue share than from making Gemini materially stickier in commerce, which strengthens its consumer AI narrative versus other models that still feel like demos. The key strategic upside is data accumulation: every shopping interaction creates intent signals that can be reused across ads, shopping, and assistant workflows, creating a flywheel that is harder for competitors to replicate than raw model quality. The risk is execution—if checkout friction, merchant availability, or recommendation quality disappoints, this becomes a feature story rather than a platform shift.

ELF is the likely relative loser on a positioning basis, not because it is directly harmed, but because ULTA’s move raises the bar for AI-enabled personalization in beauty retail. That can compress share for digitally native brands that rely on algorithmic discovery if they cannot match the convenience layer inside major platforms. Longer term, this may accelerate channel power back toward retailers and search platforms rather than brands, especially for repeat-purchase categories where recommendation quality matters more than brand equity.

The contrarian view is that the market may be overestimating how fast agentic commerce monetizes. Beauty is high-consideration, but not always high-ticket enough to justify a major behavior shift, so adoption could be strong in curiosity but modest in GMV contribution for 6-12 months. That said, even limited adoption is strategically valuable if it shifts share of search and captures first-party data, so the more important trade is relative positioning rather than absolute revenue uplift.

Key Decisions for Investors

• Long ULTA vs. short XLY basket on a 1-3 month horizon: this is a relative multiple rerating setup if investors start pricing ULTA as a commerce platform rather than a specialty retailer; downside is limited if AI traffic is merely additive, while upside comes from multiple expansion, not near-term EPS changes.
• Initiate a tactical long GOOGL position into the next 4-8 weeks: the market is still underappreciating commerce-agent optionality, and this announcement strengthens the case that Gemini can monetize beyond subscription narratives; trim if engagement metrics fail to inflect by the next quarterly readout.
• Underweight ELF relative to ULTA for 3-6 months: not a direct earnings hit, but a competitive signal that the retailer/platform layer is capturing the personalization moat; this is a good pair if you expect beauty growth to remain healthy but share to shift toward larger discovery ecosystems.
• Sell a ULTA put spread 2-3 months out if implied volatility remains elevated: monetizes the market’s tendency to overprice near-term AI headline risk while capping downside, with the thesis that the actual revenue contribution ramps slowly but sentiment stays positive.
• Set a catalyst watch for Google Shopping/AI mode adoption metrics over the next quarter: if shopper completion rates are meaningfully above low-teens conversion from trial traffic, add to GOOGL; if not, fade the move as a product demo with limited economic impact.