research-analysts2026-03-02 - 8 min

Four Tiers, One Gap: The Agentic Commerce Vendor Landscape vs. Buying Group Economics

Every existing agentic commerce vendor was built for the buyer or the individual seller. None was built for the buying group. Here is how all four existing tiers fail buying group economics — and what the fifth tier requires.

Every existing agentic commerce vendor was built for the buyer or the individual seller. None was built for the buying group.

The Fundamental Structural Problem

Every vendor currently operating in the agentic commerce space for B2B was built to serve one of two parties: the buyer or the individual seller. None was built to serve the organized buying group — the collective of independently owned distributors whose economic position depends not just on individual distributor performance, but on the aggregate volume and negotiating leverage of the network.

This is not an oversight. Buying group economics are specific to wholesale distribution. They require an architecture that simultaneously serves member distributors as individual operators and protects the network economics that make membership valuable. No horizontal proxy network has the vertical depth to build this, and no ERP vendor has the incentive to limit their product to one buying group's members.

The result is a structural gap in the vendor landscape. Here is how all four existing tiers fail the buying group — and what the fifth tier requires.

Tier 1: Enterprise Procurement AI

Vendors: Coupa, SAP Ariba, Jaggaer, Ivalua

Serves: The enterprise buyer's procurement function

Why it fails buying groups: AD's 1,400+ member distributors gain nothing from their customers using Coupa. These proxy networks optimize the buyer's procurement spend — which means helping buyers find the lowest price, best availability, and most favorable terms across the entire supplier market. "Entire supplier market" explicitly includes non-AD-member distributors. The better Coupa works for a contractor, the more it routes orders based on open-market data — the exact mechanism that erodes member volume.

Enterprise procurement AI is not a threat to buying groups; it is agnostic to buying group membership. That indifference is itself the problem.

Tier 2: ERP-Native AI

Vendors: Epicor Prism, SAP Joule, Infor Coleman, Acumatica AI

Serves: The individual distributor's internal operations

Why it fails buying groups: ERP-native AI looks inward. Demand forecasting, inventory optimization, financial analytics — these are valuable for individual distributor operations, but they do not project intelligence externally, protect buying group economics from outside agents, or enable proactive capture of autonomous agent demand.

Epicor Prism can help an AD member optimize their own branch operations. It cannot prevent an autonomous procurement agent from routing a contractor's order to a non-member based on public catalog data. It cannot alert the distributor when that routing is happening. It cannot enable the distributor to intercept that demand signal.

The ERP-native tier also faces the revenue model disruption IDC forecasts: 70% of software vendors forced to refactor away from seat-based pricing by 2028. (IDC FutureScape 2026) ERP vendors who partner with a Neutral Intelligence Proxy Layer network gain a new transaction-based revenue stream that hedges this exposure.

Tier 3: Distribution E-Commerce

Vendors: AD dataX.ai / inRiver PIM, OroCommerce, BigCommerce B2B Edition

Serves: Product discovery and e-commerce transaction execution

Why it fails buying groups: AD's digital catalog contains 7.5 million SKUs. It is an extraordinary product discovery resource — Layer 1 of the digital distribution stack. It contains zero of the 16 Immovable Values — the proprietary intelligence framework developed and owned by ProEnergy Supply that defines the relationship data embedded inside distributor ERP, CRM, and AR systems.

Product discovery tells a contractor what SKU they need. It does not tell them which member distributor has it in branch, at what price under which SPA tier, deliverable when and by which method. The 16 Immovable Values — the intelligence that actually determines which member distributor earns the transaction — live in ERP, CRM, and AR systems. Not in product catalogs.

Distribution e-commerce proxy networks solve the discovery problem. They do not solve the routing problem.

Tier 4: Emerging Agentic Proxy Networks

Vendors: Manus, Cognition, various horizontal agent orchestrators

Serves: Horizontal AI agent execution across enterprise workflows

Why it fails buying groups: Manus built remarkable technology — $100M+ ARR from orchestrating AI agent execution across enterprise tasks. Meta acquired Manus in December 2025 for $2 billion. China's National Development and Reform Commission subsequently ordered the deal unwound in April 2026. Regardless of acquisition outcome, Manus had no buying group expertise, no vertical depth in wholesale distribution, and no architecture designed around buying-group-bounded commerce.

Horizontal agents accelerate the buying group threat rather than addressing it. A more capable Manus-derived agent makes it easier for contractors to route orders through open-market AI, bypassing member distributors, SPAs, and rebate tiers in a single optimized query.

A more capable horizontal agent proxy network is not a buying group vendor. It is a more capable threat to buying group volume.

The Neutral Intelligence Proxy Layer for B2B Agentic Commerce

Proxy Network: IVAN (ProEnergy Supply LLC)

Serves: The buying group collective — simultaneously protecting member economics and enabling member growth

A Neutral Intelligence Proxy Layer network must serve both the defensive and offensive functions simultaneously — protecting existing member volume through network-bounded transactions while enabling proactive capture of autonomous agent demand from the open market.

Defensive Function. Existing member contractors order through the closed network. Their orders route to the right member distributor based on the full 16 Immovable Values — ProEnergy Supply's proprietary framework of relationship data that includes SPA pricing, credit terms, rebate tiers, territory boundaries, manufacturer authorizations, and customer history. Outside agents cannot see member data. Volume stays inside the network.

Offensive Function. IVAN delivers real-time intelligence about what autonomous agents are actively seeking in the open market. Member distributors see those demand signals and can proactively choose to engage — competing for business that would otherwise route to non-members. The member controls what data the agent sees, for which opportunity, on what terms.

ERP Enablement. IVAN shares up to 15% of its transaction fee with ERP partners who bundle IVAN as a service module. This creates a distribution partnership that addresses the ERP seat-based pricing exposure while embedding IVAN across the ERP's existing customer base. Epicor's 4,700 distribution customers represent the potential embedded distribution scale of a single partnership.

Patent-Protected Architecture. We filed for provisional patent in early 2026 and have 49 patent claims.

Protocol Fluency. IVAN speaks all four major agentic commerce protocols simultaneously: Google UCP, OpenAI ACP, Anthropic MCP, and x402. No other network in any tier operates across all four.

Exclusive Channel Access. PES holds an exclusive agreement with OpenSolar — 28,000+ solar professionals, 50,000+ designs monthly — to bring independent distributors onto the OpenSolar Shop platform, creating the first embedded B2B distributor procurement channel inside a design platform.

The Competitive Gap. No Tier 1–4 vendor is currently developing closed-network buying group capabilities with autonomous agent engagement. Analyst frameworks should evaluate whether this assessment remains accurate as the market develops — the category vacancy is the most significant unaddressed gap in the B2B agentic commerce vendor landscape.

Analyst briefings and landscape research: info@proenergysupply.com | proenergysupply.com