McKinsey Survey: 84% of European Consumers Use AI Daily -- Decision Influence Has Arrived, Execution Is Next

McKinsey Conducts First Large-Scale Survey on European Consumer AI Commerce Behavior

Europe's agentic commerce moment: Decision influence is here; execution is coming

McKinsey analysis on how agentic commerce is reshaping European retail with AI-driven decision influence already in effect.

On March 2, 2026, McKinsey & Company published a new report analyzing the current state of agentic commerce in Europe. Based on a survey of 749 consumers across France, Germany, and the United Kingdom, the report reveals how AI has already deeply penetrated the purchase "decision-making process."

The authors are Katharina Schumacher, Marcus Keutel, Philipp Kluge, and other members of McKinsey's QuantumBlack team. This is the third installment focused specifically on the European market, following the first report published in October 2025 and the automation curve paper from January 2026.

Background and Industry Trends

The agentic commerce market is moving rapidly. McKinsey predicts that by 2030, AI agents will mediate $3 to $5 trillion in global consumer commerce. Europe accounts for a significant share of global consumption, and even under conservative adoption scenarios, the impact on the retail and consumer sector cannot be ignored.

Behind this survey is the rapid development of infrastructure supporting agentic commerce. Since the start of 2026, standard protocols for agent-to-agent communication and payments have emerged in quick succession, including Google's Universal Commerce Protocol (UCP), OpenAI's Agentic Commerce Protocol (ACP), Agent2Agent (A2A), and Agent Payments Protocol (AP2). While the technical foundation is advancing, the report's value lies in providing real data on how far consumers have actually embraced AI.

Europe is also seeing distinctive regulatory developments. The EU AI Act begins enforcing strict rules from August 2026, and PSD3 (Payment Services Directive 3) is developing frameworks for AI-delegated payments. The overlapping regulatory environment of the AI Act, PSD3, GDPR, and the Consumer Rights Directive unique to Europe will significantly influence the adoption pathway of agentic commerce.

84% of European Consumers Use AI -- 38% Leverage It for Purchase Decisions

The most striking finding from the survey is that European consumer AI adoption has progressed further than expected. 84% of respondents reported using AI tools on a daily basis. By use case, general research accounts for 57%, writing and improving text for 44%, and researching products/services or making purchase decisions for 38%.

The breakdown of AI usage related to purchasing reveals even more interesting results. Comparing brands, models, prices, and reviews accounts for 63%, learning about categories and products for 55%, and discovering new products or finding inspiration for 46% -- all concentrated in the "upstream" portion of the purchase process. In other words, AI is already narrowing down options before consumers even reach a brand's website.

On the other hand, AI usage drops significantly in areas closer to "execution," such as basket building and checkout. The report analyzes this as "decision influence is expanding faster than execution infrastructure or consumer trust."

Category-level adoption rates are also revealing. While slightly higher in information-complex domains like apparel, healthcare, electronics, and travel, the majority of categories cluster in the 30-36% range. AI is beginning to function not just in specific specialized categories but as a general-purpose "decision support layer." The sole exception is eyewear (13%), where the added value of current AI tools remains limited in domains requiring physical try-on and in-person evaluation.

Consumer Trust Centers on "Judgment," Remains Cautious on "Autonomous Action"

Another critical finding from the survey is the boundary of trust. Consumers place high trust in "decision support" functions such as review summarization, trade-off presentation, option comparison, and optimal choice recommendations. However, trust drops sharply when it comes to "action delegation" -- automatic basket building, checkout completion, and recurring automatic orders.

McKinsey interprets this trend as "consumers are not rejecting AI's judgment -- they are rejecting unrestricted delegation of authority." The three conditions for high trust are: "the action is reversible," "accountability is clear," and "explicit consent is given." This is not European consumer conservatism but rather carries the positive meaning that consumers are defining the conditions of delegation themselves.

Cross-Border Commerce Europe has also noted this shift, arguing that marketers need to design "marketing aimed at both humans and machines." In an era where consumer AI agents become the first customer touchpoint, traditional ad placement and UI optimization alone are becoming insufficient.

Impact and Actions for E-Commerce Businesses

In this report, McKinsey presents three strategic actions for e-commerce businesses and brands.

First, become "chosen by AI." As discovery and comparison shift to AI interfaces, competitive advantage depends not only on being recognized by consumers but on whether AI agents can access and understand your offerings. Structured product metadata, consistent naming and taxonomy, evidence-based claims, and third-party trust signals such as reviews and expert evaluations become prerequisites for appearing in the AI journey. McKinsey describes this as "Agent Engine Optimization," positioning it as a new optimization domain replacing SEO. Adapting to interoperability standards like MCP (Model Context Protocol) and UCP is also urgent.

Second, differentiate through "explainability." In the agentic commerce world, clarity, comparability, and evidence become sources of differentiation over persuasive UX. "What is the product," "who is it for," "what are the trade-offs," and "why does it fit this context" need to be structured in ways that communicate to both humans and AI.

Third, design for participation in the "agent web." Whether to build your own consumer-facing AI agent is a discretionary strategic decision, but being ready to participate in the agent network -- the coordination between personal agents, retail agents, and broker agents -- is essential. Protocols like ACP, A2A, AP2, and UCP enable inter-agent discovery and negotiation while maintaining a design where retailers remain the merchant of record, retaining control over identity, payment authentication, policies, fulfillment, and post-purchase services.

Conclusion

McKinsey's latest report demonstrates with real data that agentic commerce in Europe is progressing in a two-tier structure: "widespread adoption of AI-driven decision support" and "cautious transition to autonomous execution." The figures showing 84% of consumers using AI daily and 38% leveraging it for purchase decisions mean that agentic commerce is no longer a future story but a change happening in real time.

European consumers' "caution" is not a weakness but could serve as a model case for trust-based agentic commerce. The European model of "conditional delegation" premised on reversibility, accountability, and explicit consent aligns well with the regulatory frameworks of the EU AI Act and PSD3.

For e-commerce businesses, the next developments to watch are the EU AI Act enforcement in August 2026 and the expanding implementation of protocols like UCP and ACP. Whether businesses can build data foundations that allow AI agents to "discover" them starting now will determine their future competitiveness.