INTERMEDIARY RETAINED
AI discovery→Marketplace→Provider service
A new route into the existing intermediary.
The marketplace can gain distribution while keeping its supply, account and completion role.
COMMERCE / WHO GAINS?
Who gains from AI commerce partnerships?
Companies gain leverage when partners need their contribution and cannot readily replace it. Reaching more customers or retaining fulfilment does not, by itself, establish that advantage.
Compare who gains and who becomes exposed ↓THE COMPETITIVE LANDSCAPE
Who gains under different partnership arrangements?
A firm can gain distribution while losing influence over customer choice. Compare who gains reach, who faces substitution and which conditions change the outcome.[1]
Potential shifts in competitive position. Both parties can gain.
01 / HOW ARRANGEMENTS DISTRIBUTE ADVANTAGE
How rebundling reinforces or bypasses intermediaries
The same AI capability can reinforce an intermediary—or bypass it. Moving discovery into AI does not determine the winner. Where the customer goes next, and which services that destination supplies, change the competitive outcome.[2]
DIRECT PROVIDER ROUTE
AI discovery→Provider completion
Intermediary outside this route
A possible route around the intermediary.
Bypass matters only if the new connection replaces the service and attracts transactions that would otherwise use it.
Compare how customer choice changes competitive position
| The customer wants… | The channel’s position | The provider’s position |
|---|---|---|
| A product from any equivalent seller | Can direct the order between competing sellers. | Can gain sales while losing a reason for the customer to return directly. |
| A particular provider’s assortment, benefits or service | Gains a useful offering it cannot replace with just any seller. | Can extend its reach while retaining demand for its own service. |
| A service previously reached through an intermediary | Can connect to that intermediary—or create a direct provider route. | Gains a bypass opportunity only if the direct route delivers the service customers need. |
A platform can gain distribution while merchants behind it become less distinctive.
02 / WHO DEPENDS ON WHOM?
Customer access, service capabilities and bargaining power
Leverage tends to favour the party with better alternatives for the objective at stake. A provider’s position can strengthen when the channel needs its contribution more than the provider needs that particular channel. If both have alternatives—or both need each other—the balance differs. The strategic comparison is between each party's ability to replace the contribution and reach the customers it needs.[3]
PROVIDER’S ALTERNATIVE CUSTOMER ROUTES →
Several viable routesDependent on this route
03 / HOW ADVANTAGE SHIFTS
How standards and adoption change advantage
As more merchants join an AI channel, customers gain more options and merchants face more substitutes. Merchants gain alternatives too when they can use the same connections across competing channels. Accounts and integrations provide alternatives only when they are portable and commercially usable.[4]
Explore how advantage shifts →
COMPARE THE DIRECTION OF CHANGE
When does AI unbundle the customer relationship, and when does it strengthen the existing business?
An AI channel can separate product discovery from the provider's own experience. An internal assistant brings AI capability into that experience instead. Shopify's ChatGPT referral arrangement and Lowe's own assistant illustrate these different directions. The first introduces a new intermediary in customer choice; the second introduces a capability supplier inside the business. Neither arrangement, by itself, establishes stronger customer relationships or greater independence.[5]
Shopify merchants through ChatGPT
Customer → ChatGPT discovery → merchant store
In the March 2026 referral configuration, discovery starts outside the merchant storefront. Shopify connects the catalogue and store infrastructure; the merchant completes the sale.
The route into the business changes, creating another way for customers to reach it.
Lowe’s Mylow on Lowes.com
Customer → Lowe’s advisor → Lowe’s product guidance
OpenAI models power project advice and product recommendations within Lowe’s own website. Mylow Companion also supports store associates.
What changes: the capability supplied inside the existing destination. The described experience does not require a separate lab shopping channel.
The comparison contrasts an external customer channel with AI inside the retailer's service. The Lowe’s account comes from its model supplier.
WHAT THIS MEANS FOR YOUR BUSINESS
Choose a partnership for the position you want to build
Start with your objective and the dependency you can accept. These choices can combine: a retailer can run its own assistant and join AI channels; a platform can provide the route into those channels.
Use these as conditional options, not company-specific recommendations. Other partnership purposes—including rights, joint development and infrastructure—need further commerce-specific assessment.
APPLY THIS TO YOUR POSITION
Bring your objective and current partnerships. Advisory work tests these options against your capabilities, evidence and terms.
Apply this to your business.
Explore advisory →Research and interpretation
The framework draws on 81 dated episode/role records across 58 direct commerce case bundles. The records distinguish plans, historical episodes and unresolved allocations.
Sources and notes
Case evidence. Partnerships Observatory, reconciled Commerce research, cutoff 7 October 2026: 81 dated episode/role records across 58 direct case bundles, including historical, planned and specification material. The displayed record disclosures retain the original source passages, dates and unresolved allocations.
Third-party analysis. McKinsey, "The agentic commerce opportunity: How AI agents are ushering in a new era for consumers and merchants" (17 October 2025), chapters 1-2, direct and brokered interaction models and interoperability. These distinctions inform the explanation; the Observatory applies them to the arrangements and capabilities in its research.
Theory. Richard M. Emerson, "Power-Dependence Relations," American Sociological Review 27, no. 1 (1962), pp. 31-41, especially pp. 32-33. Dependence concerns valued outcomes and costly alternatives for the same outcome. The strategic application here is the Observatory's; the analysis examines how alternatives affect negotiating room.
Theory. Carliss Y. Baldwin and Marcus Holgersson, "Ecosystem Evolution: Contracts, Modules, and Standards," forthcoming handbook chapter, held manuscript, pp. 7-11. Compatible modules and design rules can permit replacement. The commerce application remains conditional on working compatibility and usable participation rights.
Case evidence. Partnerships Observatory, corrected own-surface-assistant findings, cutoff 7 October 2026, supplier-route and source-argument comparisons covering Lowe's, Macy's, Ulta and other named cases. Model supply, packaged agents and cloud/vendor mediation are distinguished. Interface retention is observed where stated; data rights, practical switching and causal performance gains remain unestablished. Source attributions and dated case disclosures remain in the research.
Theory. Michael G. Jacobides and Sidney G. Winter, "Capabilities, Transaction Costs, and Evolution: Understanding the Institutional Structure of Production" (July 2003 working paper), opening argument on relative capabilities and firm boundaries. Capability sourcing and the division of work are separate from a demonstrated competitive advantage. The firm-specific interpretation here is the Observatory's.
Third-party analysis. Boston Consulting Group, "Agentic Commerce Is Redefining Retail - How to Respond" (2025), "The Threats to Retailers" and retailer-response sections. Its analysis of AI-mediated choice and customer contact informs these questions.