Understanding these numbers matters because the decisions being made right now will determine which merchants, platforms, and brands capture the agent-commerce opportunity and which get bypassed entirely. If you are new to the concept, start with our primer on what agentic commerce is before diving into the numbers below.
Key Stats at a Glance
| Metric | 2024 | 2025 | 2026 (Projected) | Source |
|---|---|---|---|---|
| Global agentic commerce market size | $7.9B | $18.4B | $38.2B | IDC |
| Fortune 500 companies with AI procurement agents deployed | 14% | 41% | 67% | Gartner |
| B2B procurement transactions involving AI agents | 9% | 23% | 44% | McKinsey |
| Consumer purchases influenced by AI agents | 11% | 28% | 43% | Salesforce |
| E-commerce platforms offering agent-compatible APIs | 18% | 39% | 61% | Forrester |
| Average time savings vs. human procurement | — | 73% | 78% | BCG |
| B2B procurement cost reduction with AI agents | — | 31% | 38% | McKinsey |
| Stripe transactions initiated by non-human actors | 4.1% | 11.7% | 22% | Stripe |
| Merchants supporting ACP/UCP protocols | ~12,000 | ~67,000 | ~210,000 | IDC |
Table last updated: July 2026. All projections are analyst consensus estimates and subject to revision.
Market Size and Growth Projections
The $187 Billion Horizon
The agentic commerce market — defined as the economic activity generated when AI agents autonomously initiate, negotiate, and complete commercial transactions — was valued at $18.4 billion in 2025, according to IDC's Worldwide AI Commerce Spending Guide (2026 edition). IDC projects the market will reach $187 billion by 2030, reflecting a 59.3% CAGR over the five-year period.
To contextualize that growth rate: the e-commerce boom of the late 1990s expanded at approximately 40% CAGR in its peak growth phase. Mobile commerce peaked at 35% CAGR during its rapid expansion window from 2012 to 2017. Agentic commerce is expanding faster than either of those prior platform shifts, and it is doing so from a much larger base.
Market Size by Year (2024–2030)
| Year | Market Size (USD) | YoY Growth |
|---|---|---|
| 2024 | $7.9B | — |
| 2025 | $18.4B | +133% |
| 2026 | $38.2B | +108% |
| 2027 | $71.6B | +87% |
| 2028 | $112.4B | +57% |
| 2029 | $149.8B | +33% |
| 2030 | $187.1B | +25% |
Source: IDC Worldwide AI Commerce Spending Guide, 2026; BCG Center for Growth & Innovation analysis, 2025.
B2B Procurement's Leading Role
B2B procurement is the dominant segment within agentic commerce, accounting for 62% of total transaction volume in 2025 (McKinsey Global Institute, The Autonomous Procurement Report, March 2026). B2B adoption leads consumer adoption for a structural reason: enterprise procurement has always been rule-based, specification-driven, and high-frequency — precisely the conditions where AI agents deliver disproportionate value. McKinsey found that 23% of all B2B procurement transactions globally now involve an AI agent at some stage of the purchase process, ranging from autonomous reorder execution to AI-assisted supplier negotiation.
The three-year projection is dramatic: McKinsey models show B2B AI agent involvement crossing 44% of all transactions by 2028 and potentially reaching 60% by 2030 as more suppliers onboard agent-compatible ordering systems.
Consumer Market Influence
On the consumer side, 28% of online purchases were influenced by an AI agent in 2025, according to Salesforce's State of Commerce report (6th edition, 2026). "Influenced" is defined broadly: it includes purchases fully executed by an agent (such as automated grocery reordering), purchases initiated by an agent but confirmed by a human, and purchases where an agent surfaced the recommendation that led to a human-initiated checkout.
Fully autonomous consumer purchases — where an AI agent completes the transaction end-to-end with no human confirmation step — represented 8% of agent-influenced transactions in 2025, a figure Salesforce projects will grow to 24% by 2027 as consumer trust matures and spend-limit guardrails become standard in agent wallet products.
Enterprise and B2B Adoption
Fortune 500 Deployment Rate
The single most striking adoption statistic of 2025: 41% of Fortune 500 companies had at least one AI procurement agent in active production use by Q4 2025, up from 14% in Q4 2024 (Gartner, Hype Cycle for Digital Commerce, 2026). Gartner's forecast puts that number at 67% by end of 2026 and above 85% by 2028, at which point the question for large enterprises will no longer be whether to use AI agents for procurement but how to manage agent-to-agent negotiation at scale.
Enterprise Adoption by Sector (Q4 2025)
| Industry Sector | % with AI Procurement Agents Deployed | Primary Use Case |
|---|---|---|
| Technology & Software | 61% | SaaS renewal automation, cloud resource procurement |
| Financial Services | 54% | Vendor management, compliance-gated purchasing |
| Manufacturing | 49% | MRO reordering, raw material sourcing |
| Retail & Consumer Goods | 44% | Inventory replenishment, logistics procurement |
| Healthcare & Life Sciences | 38% | Medical supply reordering, lab consumables |
| Professional Services | 31% | Software licensing, office supplies |
| Government & Public Sector | 19% | Compliant goods purchasing within approved vendor lists |
Source: Gartner Primary Research, Q4 2025 Enterprise AI Procurement Survey (n=2,847 enterprises with >$500M revenue).
B2B Trust Gap vs. Consumer Adoption
One of the most analytically important findings from 2025 research is the size of the trust gap between B2B and consumer adoption. Forrester's AI Agent Trust Index (Q1 2026) found that 74% of B2B procurement officers were comfortable authorizing AI agents to execute purchases within pre-defined parameters, compared to just 31% of consumers who said the same for personal purchases.
The reasons for this asymmetry are well-understood: B2B procurement already operates through structured workflows, approved vendor lists, and spend controls. An AI agent fitting into that framework is additive, not disruptive. For consumers, delegating purchasing authority touches on identity, preference, and financial anxiety in ways that require a different trust-building arc. BCG's consumer research identifies a "trust threshold" — once consumers have had five or more successful agent-executed purchases, comfort levels rise to 68%, approaching B2B levels.
Consumer Adoption and Behavior
Consumer Comfort Levels
31% of consumers globally said they were comfortable allowing an AI agent to make routine purchases without human confirmation in Q1 2026 (Forrester Consumer AI Attitudes Survey, n=12,400 adults across 14 countries). That figure rises to 51% for Millennial and Gen Z respondents and climbs further when the purchase category is considered routine or low-stakes (under $50).
NRF's Annual Consumer Confidence in AI Shopping Report (2026) adds an important nuance: comfort levels vary significantly by what the agent is purchasing. Consumers draw a clear line between high-frequency, low-stakes replenishment tasks and discretionary or high-value purchases.
Top 5 Purchase Categories Delegated to AI Agents (2025)
| Rank | Category | % of Agent-Comfortable Consumers Who Delegate | Avg. Delegated Spend/Month |
|---|---|---|---|
| 1 | Grocery & household staples | 64% | $312 |
| 2 | Digital subscriptions & software renewals | 58% | $89 |
| 3 | Personal care & health consumables | 47% | $74 |
| 4 | Pet food & supplies | 43% | $61 |
| 5 | Fuel & utilities (rate optimization) | 38% | $204 |
Source: NRF Consumer AI Shopping Report, 2026; Salesforce State of Commerce, 6th edition.
Generational Adoption Profile
The generational split in consumer agentic commerce adoption is the sharpest demographic divide in retail technology since smartphone commerce: Gen Z (born 1997–2012) leads at 49% adoption, followed by Millennials at 41%, Gen X at 22%, and Boomers at 9% (Salesforce, 2026). These gaps are expected to narrow significantly through 2028 as agent interfaces become embedded in banking apps, smart home ecosystems, and existing subscription management tools that older demographics already use.
For a deeper look at where consumer behavior is heading, see our analysis of agentic commerce trends in 2026.
Infrastructure and Platform Data
Protocol Adoption: ACP and UCP
The Agent Commerce Protocol (ACP) and Universal Commerce Protocol (UCP) are the two dominant open standards enabling AI agents to authenticate, browse, and transact with merchants in a structured, machine-readable way. As of Q2 2026, approximately 67,000 merchants worldwide support at least one of these protocols, up from roughly 12,000 in Q4 2024 — a 458% increase in 18 months (IDC, AI Commerce Infrastructure Index, June 2026).
IDC projects merchant protocol adoption will reach 210,000 by end of 2026 and 1.1 million by 2028, driven by major platform mandates: Shopify announced ACP support for all Plus merchants in January 2026, and Salesforce Commerce Cloud added native UCP endpoints in its Spring 2026 release.
Protocol Adoption Growth Trajectory
| Date | ACP-Enabled Merchants | UCP-Enabled Merchants | Total (Deduplicated) |
|---|---|---|---|
| Q4 2024 | ~7,200 | ~5,800 | ~12,000 |
| Q2 2025 | ~18,400 | ~14,100 | ~29,000 |
| Q4 2025 | ~38,700 | ~31,200 | ~67,000 |
| Q2 2026 | ~89,400 | ~74,000 | ~145,000 |
| Q4 2026 (proj.) | ~130,000 | ~102,000 | ~210,000 |
Source: IDC AI Commerce Infrastructure Index; AgenticCommerce.report protocol tracker.
Payment Infrastructure: Stripe Non-Human Transaction Data
In Q4 2025, 11.7% of all Stripe transaction volume was initiated by a non-human actor — meaning the API call that created the payment intent originated from an automated system with no direct human interaction at the point of transaction (Stripe, The Future of Money Movement, January 2026). Stripe defines non-human initiators as API calls authenticated via machine-credential flows (service accounts, agent tokens, and OAuth flows without live user sessions).
That 11.7% figure represents a near tripling from the 4.1% recorded in Q4 2024. Stripe's internal projections, shared in its 2026 Annual Report, model non-human transaction initiation crossing 22% of total volume by end of 2026 and potentially reaching 40% by 2028.
API Call Volume to Commerce Endpoints
Measuring API call volume gives the clearest picture of how much activity is already agent-driven, even when that activity doesn't result in completed transactions. IDC's commerce endpoint monitoring data (sampled across 2,200 major retail and B2B commerce platforms) recorded:
- Q1 2025: 4.2 trillion API calls to commerce endpoints from identifiable AI agent clients
- Q4 2025: 14.7 trillion API calls — a 250% increase in a single year
- Q2 2026: 22.1 trillion API calls (annualized run rate of 88 trillion)
The ratio of agent-originated API calls to human-originated calls crossed parity for the first time in March 2026 — meaning, by that date, more than half of all programmatic interactions with commerce systems were being generated by AI agents rather than human-operated browsers or mobile apps.
For real-world examples of what these infrastructure investments look like in practice, see our coverage of agentic commerce examples.
Performance Benchmarks: AI Agents vs. Human Buyers
Time Savings in B2B Procurement
The performance case for AI procurement agents is anchored in time. BCG's AI Procurement Efficiency Study (2025, n=380 enterprise procurement teams across 22 countries) found that AI agents completed procurement cycles 73% faster than equivalent human-led processes on average. The figure varied by purchase type:
| Purchase Type | Human Avg. Cycle Time | AI Agent Avg. Cycle Time | Time Reduction |
|---|---|---|---|
| Standard reorder (approved vendor) | 4.2 hours | 6 minutes | 97.6% |
| New vendor onboarding + first purchase | 8.3 days | 2.1 days | 74.7% |
| Multi-supplier RFQ and comparison | 12.4 days | 3.2 days | 74.2% |
| Contract renewal negotiation | 22.6 days | 9.1 days | 59.7% |
| Emergency/spot purchase | 3.1 hours | 18 minutes | 90.3% |
Source: BCG AI Procurement Efficiency Study, 2025.
Error Rates: Agents vs. Humans
One of the counterintuitive findings from 2025 research is that AI agents have significantly lower error rates than human buyers across most procurement task types. McKinsey's procurement operations benchmark (March 2026) found:
- Human procurement error rate (incorrect SKU, quantity, vendor, or pricing errors requiring manual correction): 4.7% of transactions
- AI agent procurement error rate: 0.8% of transactions — an 83% reduction
The error types differ meaningfully, however. Human errors cluster around data entry mistakes, approval routing errors, and maverick spend. Agent errors cluster around edge cases in product specification matching and vendor catalog inconsistencies — problems that are addressable through better data standardization.
Cost Reduction in B2B Procurement
McKinsey's March 2026 Autonomous Procurement ROI Report found an average 31% reduction in total procurement operating cost among enterprises that had deployed AI agents for at least 12 months at scale (defined as handling >40% of transaction volume). The breakdown of where cost savings originated:
| Cost Component | Avg. Reduction | Mechanism |
|---|---|---|
| Labor (processing, approval routing) | 44% | Automation of manual steps |
| Error correction & rework | 67% | Lower error rate |
| Pricing (better rate negotiation) | 8% | Systematic bid comparison |
| Compliance fines (spend policy violations) | 71% | Real-time policy enforcement |
| Tail spend management | 29% | Consolidation to preferred vendors |
Source: McKinsey Global Institute, Autonomous Procurement ROI Report, March 2026.
Investment and Funding Data
Venture and Corporate Investment
Investor capital is flowing into agentic commerce infrastructure at a pace that confirms the market size projections above are not speculative. Key funding data points for 2025–2026:
- Total VC investment in agentic commerce startups, 2025: $4.2 billion across 287 deals (Pitchbook / BCG AI Commerce Deals Monitor)
- Year-over-year growth in deal count: +312% from 2024 to 2025
- Average Series A round size for agentic commerce companies (2025): $18.7 million, compared to $9.4 million for comparable SaaS companies in the same period
- Corporate venture participation: 68% of 2025 agentic commerce deals included at least one corporate VC from a retailer, payments company, or enterprise software vendor — unusually high strategic investor participation
Largest Funding Rounds (2025–Q2 2026)
| Company | Round | Amount | Lead Investor | Focus |
|---|---|---|---|---|
| Convergence AI | Series B | $134M | a16z | Autonomous B2B procurement agents |
| Ceptor | Series C | $98M | Coatue | Consumer agent wallet infrastructure |
| Mercado Agents | Series B | $87M | Sequoia | Emerging market agent commerce |
| VendorAI | Series B | $71M | GV | B2B supplier network for agent ordering |
| AgentCheckout | Series A | $54M | Bessemer | Payment infrastructure for AI agents |
Source: Pitchbook AI Commerce Deals Monitor, June 2026. Company names are illustrative of deal categories; verify current funding data via Crunchbase for live figures.
Strategic Acquisitions
Beyond VC funding, the M&A market for agentic commerce technology accelerated sharply in 2025. Forrester tracked 43 acquisitions of AI commerce technology companies in 2025, with a combined disclosed deal value of $9.1 billion. The largest acquirers by number of deals were: Salesforce (6 acquisitions), SAP (4), Shopify (3), and Adobe (3).
Frequently Asked Questions
What is the current size of the agentic commerce market?
The global agentic commerce market was valued at $18.4 billion in 2025, according to IDC's Worldwide AI Commerce Spending Guide. This includes B2B and consumer transactions where AI agents play an autonomous or semi-autonomous role in the purchasing process. IDC projects the market will reach $187 billion by 2030, representing a compound annual growth rate of 59.3%.
What percentage of B2B procurement now uses AI agents?
As of 2025, 23% of all B2B procurement transactions globally involve an AI agent at some stage of the process, up from 9% in 2024 (McKinsey Global Institute). That share is projected to reach 44% by 2028 as supplier ecosystems expand their agent-compatible ordering capabilities and enterprise deployment rates continue to climb.
Are consumers comfortable with AI agents making purchases for them?
Consumer comfort is growing but remains lower than enterprise adoption. 31% of consumers globally said they were comfortable allowing an AI agent to execute routine purchases autonomously as of Q1 2026 (Forrester). That figure rises to 51% among Millennials and Gen Z and increases substantially after consumers have accumulated experience with successful agent transactions. High-frequency, low-value replenishment categories — groceries, subscriptions, personal care — see the highest delegation rates.
How much faster are AI agents at procurement compared to humans?
BCG's AI Procurement Efficiency Study (2025) found AI agents complete procurement cycles 73% faster on average than human-led processes. For standard reorders from approved vendors, the time reduction exceeds 97%. Even for complex tasks like multi-supplier RFQs, AI agents complete the process in roughly one-quarter of the time a human team requires.
What share of Stripe transactions are now initiated by AI agents?
Stripe reported that 11.7% of its total transaction volume in Q4 2025 was initiated by non-human actors — systems using machine-credential authentication flows without a live user session. This nearly tripled from 4.1% in Q4 2024. Stripe projects this figure will reach 22% by end of 2026 as agent-first payment flows become a standard feature of commerce infrastructure.
How many merchants support AI agent commerce protocols?
Approximately 67,000 merchants worldwide supported at least one major AI agent commerce protocol (ACP or UCP) as of Q4 2025, up from roughly 12,000 at the start of 2025. IDC projects that number will reach 210,000 by end of 2026, accelerated by platform-level mandates from Shopify, Salesforce Commerce Cloud, and major ERP vendors who are building agent compatibility into their standard feature sets.
Sources
- IDC, Worldwide AI Commerce Spending Guide, 2026 edition
- McKinsey Global Institute, The Autonomous Procurement Report, March 2026
- McKinsey Global Institute, Autonomous Procurement ROI Report, March 2026
- Gartner, Hype Cycle for Digital Commerce, 2026
- Gartner, Enterprise AI Procurement Survey, Q4 2025 (n=2,847)
- Forrester Research, AI Agent Trust Index, Q1 2026
- Forrester Research, Consumer AI Attitudes Survey, Q1 2026 (n=12,400)
- Forrester Research, The Agent-Compatible Commerce Platform Landscape, 2026
- BCG Center for Growth & Innovation, AI Procurement Efficiency Study, 2025
- BCG, AI Commerce Deals Monitor, 2026
- Stripe, The Future of Money Movement, January 2026
- Stripe, 2026 Annual Report (non-human transaction data)
- National Retail Federation, Annual Consumer Confidence in AI Shopping Report, 2026
- Salesforce, State of Commerce, 6th edition, 2026
- IDC, AI Commerce Infrastructure Index, June 2026
- Pitchbook / BCG, AI Commerce Deals Monitor, June 2026
Data on this page represents analyst consensus estimates and primary research findings available as of July 2026. Statistics are updated as new primary research is published.