Recent Posts

AI agents

When the Buyer Is an AI Agent – How Georgian Business Needs to Change

Key Takeaway The next phase of e-commerce may be defined by a simple change: consumers will increasingly delegate product

When the Buyer Is an AI Agent – How Georgian Business Needs to Change

Key Takeaway

The next phase of e-commerce may be defined by a simple change: consumers will increasingly delegate product discovery, comparison, and sometimes purchasing to AI agents. Instead of manually searching dozens of pages, a customer can give an agent a goal, budget and constraints and let the system evaluate the market.

A 2026 thought-leadership collaboration between the Yale Center for Customer Insights and Boston Consulting Group argues that brands increasingly need to compete for two decision makers: people and the AI systems that help shape their choices.

According to BTU researchers, this could be unusually important for Georgian SMEs. They may not be able to outspend global brands in advertising, but they can compete through cleaner product data, transparent pricing, credible third-party evidence and reliable operational delivery.

Agentic Commerce Begins When AI Moves from Advice to Action

Consumers already use generative AI to ask which phone to buy, where to stay or which product best fits a need. Agentic commerce goes further. An agent can remember criteria, compare multiple sources, check availability and pricing, prepare an order and, with the right authorization, increasingly participate in the transaction itself.

This is no longer only a forecast. Google’s Universal Commerce Protocol, introduced for the emerging agentic-commerce environment, is designed as a common language for platforms, agents and merchants, supporting product discovery, catalog search and checkout workflows.

McKinsey estimates that AI agents could mediate between USD 3 trillion and USD 5 trillion of global consumer commerce by 2030. That is a forecast, not a guaranteed outcome, but it explains why commerce infrastructure is already being redesigned.

Companies Will Need to Serve Two Decision Makers

Human buyers respond to brand, design, identity, experience and emotion. AI agents operate differently. They can compare price, availability, specifications, shipping, return policies, reviews and other observable signals at scale.

That does not make AI perfectly objective. Agents depend on user instructions, available data and the systems through which they operate. Bad or incomplete data can produce bad choices.

But the strategic implication remains: persuasive claims increasingly need machine-checkable evidence.

Brand Promise Must Match Brand Performance

BCG distinguishes between brand promise-what the company says-and brand performance-what customers actually experience.

If a brand promises speed, AI systems can compare delivery outcomes and complaints. If it promises transparent pricing, systems can examine final prices and extra fees. If a brand promises easy returns, the policy and customer feedback can be evaluated.

In agentic commerce, operational inconsistency becomes a marketing problem because AI can aggregate evidence across many sources more quickly than a typical customer.

SEO Is Not Disappearing, but Machine-Readable Commerce Is Becoming More Important

Generative Engine Optimization, or GEO, has become a popular label. Google’s 2026 guidance is important here: from Google’s perspective, optimizing for generative AI search remains part of good search optimization rather than a completely separate replacement for SEO.

What is changing is the importance of structured, explicit commerce data. Google’s merchant documentation supports product data covering price, availability, shipping, returns and reviews.

BTUAI groups the practical readiness challenge into four working areas. This is an analytical synthesis, not an official Yale/BCG four-pillar standard:

  • Entity clarity: consistent identity and product facts across channels;
  • Extractability: structured product data, clear HTML, tables and APIs;
  • Trust signals: independent reviews, credible citations and transparent policies;
  • Freshness: continuously updated price, inventory, delivery and policy data.

Georgia Is Already Digital Enough for the Shift to Matter

Geostat data show that in June 2025, 37.4% of internet users aged 15 and older had bought or ordered goods or services online during the previous 12 months. The figure was 31.3% one year earlier.

BTUAI’s calculation shows an increase of 6.1 percentage points, or about 19.6% relative to the 2024 level. Among 15–29-year-olds, the online purchasing rate was already 61.1%.

These figures do not measure agentic commerce. They show that online purchasing is already an established consumer behavior in Georgia, especially among younger users. If product discovery moves toward AI interfaces, Georgian commerce will be exposed to that shift.

B2B May Move Faster Toward Machine Buyers

Procurement is often a natural environment for buyer agents because criteria are more explicit: price, quantity, technical specifications, delivery time, contract terms and supplier reliability.

One source claim needs correction. Forrester’s official 2026 prediction does not support the idea that 90% of B2B purchases will be agent-to-agent.

Forrester instead predicts that 20% of B2B sellers will be forced to engage in agent-led quote negotiations in 2026. Its research also says 61% of purchase influencers report that their organizations have or will use a private generative AI engine to support purchasing.

The direction is clear, but forecasts should not be confused with current adoption.

CMOs See the Transformation, but Most Have Not Reached Agentic Execution

BCG’s 2026 survey of nearly 300 global CMOs found that 96% believe AI is driving end-to-end transformation of marketing. Yet only 31% had made agentic execution a reality.

The gap is 65 percentage points. Awareness is far ahead of operational implementation.

For Georgian firms, that gap creates an opening. A smaller company can become AI-ready through disciplined data and operations without matching the media budgets of multinational competitors.

Illustrative Georgian Mini-Case: Selling Georgian Wine Online

Imagine a small Georgian winery selling to international customers. Its website is attractive but each product page relies mostly on emotional phrases: unique Georgian taste, tradition, premium quality.

A human customer may respond to the story. But an AI agent looking for a dry Saperavi under EUR 25, with a specific vintage, shipping to the EU and a clear return policy needs facts.

If a competitor provides precise SKU data, price, inventory, origin, alcohol content, delivery time, reviews and updated structured product information, the agent may rank that competitor more highly even if the Georgian producer has a stronger story.

The answer is not to remove emotion from branding. It is to keep the story for humans and provide the evidence for machines.

What Georgian Businesses Should Do Now

  • Audit how major AI assistants describe and compare the brand and identify missing or incorrect facts.
  • Create one authoritative product-data source used consistently across channels.
  • Implement relevant product, offer, shipping, return and organization structured data.
  • Treat GEO as an extension of strong search and content practices, not a replacement for SEO.
  • Connect marketing with operations so brand promise matches delivery performance.
  • Build independent trust signals through reviews, earned media and credible references.
  • Keep price, inventory, delivery and policy information fresh.
  • Prepare for APIs and emerging commerce protocols, especially in online export and B2B.

BTU Researchers’ Assessment

According to BTU researchers, the most important change in agentic commerce is not that AI replaces the customer. It is that a new intermediary appears between the customer and the market-one capable of comparing evidence across hundreds of alternatives in seconds.

Marketing remains important, but its character changes. Human trust still depends on story and experience, while machine-mediated trust increasingly depends on accessible evidence.

For Georgia, the strongest response may be speed and clarity rather than advertising scale: clean data, transparent prices, current information and reliable fulfillment.

Key Findings

  • Agentic commerce describes a market in which AI agents participate more actively in product discovery, evaluation and increasingly transactions.
  • McKinsey estimates that AI agents could mediate USD 3–5 trillion of global consumer commerce by 2030.
  • BCG and Yale argue that companies increasingly need to consider both human decision makers and AI-mediated decision pathways.
  • In BCG’s survey of nearly 300 CMOs, 96% see end-to-end AI transformation, but only 31% have made agentic execution a reality.
  • Forrester’s official 2026 forecast does not support a 90% agent-to-agent B2B purchasing claim; it forecasts agent-led quote negotiations affecting 20% of B2B sellers.
  • In Georgia, 37.4% of internet users aged 15+ bought or ordered online in 2025; the rate was 61.1% among 15–29-year-olds.
  • For Georgian companies, machine-readable product information and real operational credibility may become new sources of competitive advantage.

Why This Matters for Georgia

Agentic commerce could matter most in Georgia’s export-oriented SMEs, tourism, hospitality, financial services, e-commerce and B2B supply. Small firms often reach foreign customers almost entirely through digital channels.

If AI increasingly mediates discovery and comparison, product-data quality becomes part of competitiveness. This is a strategically important opportunity that does not necessarily require a large advertising budget.

Conclusion

The human buyer is not disappearing. But a second buyer is emerging alongside the human: the AI agent.

Businesses will therefore need to earn two forms of trust-human trust in the brand and machine-readable trust in the evidence.

For Georgian companies, the next marketing question will not only be ‘How do we get the customer’s attention?’ It will also be: ‘Can the agent understand exactly what we sell, what it costs, how reliably we deliver, and why we should be selected over a competitor?’

Data and Main Sources

  • Boston Consulting Group & Yale Center for Customer Insights – Agentic AI Will Make the CMO’s Role More Consequential, 2026.
  • McKinsey & Company – Agentic Commerce research, 2025–2026.
  • Forrester – 2026 B2B Marketing, Sales, and Product Predictions.
  • Google Developers – Universal Commerce Protocol; product structured data; AI search optimization guidance.
  • National Statistics Office of Georgia – ICT Usage in Households, e-commerce indicator, June 2025.
  • BTUAI Research Team – Georgia-focused analytical interpretation.

Prepared by the academic team of Business and Technology University and the BTUAI Research Team, Tbilisi, Georgia.