Where to participate
Prioritise channels that attract relevant customers without eliminating practical alternatives.
COMMERCE / WHO GAINS? / HOW ADVANTAGE SHIFTS
Which investments stay valuable as AI commerce partnerships change?
A connection can become easier to copy or cease to operate as a channel changes. A commerce business has a more durable position when customers and partners still need its service across several arrangements.
Explore what changes the advantage ↓01 / STANDARDISATION
The change as connections standardise
Common interfaces can make it easier for a commerce platform to connect merchants to AI channels while reducing the need for bespoke integration.[1][2] Shopify can therefore gain reach as some connection work becomes less distinctive. Its merchant access and ability to support purchases are separate contributions.[3] This shift depends on working compatibility across implementations; publishing a standard does not make every connection interchangeable.
02 / CONCENTRATION AND CONTINUITY
How demand and alternatives change the balance
An AI channel gains negotiating room if shoppers keep returning and accept several sellers as substitutes. Providers gain alternatives if the connections, accounts and operating experience they develop can be used in other channels. Both effects can occur as the ecosystem grows. Compare the demand a provider would lose by leaving with the service the channel would lose without it.[4]
03 / ADOPTION AND REVERSAL
Strategic implications for investment
A commerce business can separate investment in its underlying service from investment tied to a particular channel design. The Shopify checkout sequence below shows why: discovery through ChatGPT continued through a different checkout route after the specified hosted route was withdrawn. Work tied to the original checkout faced a change in use, while the wider relationship continued. The sequence does not show that every embedded app ended or establish why OpenAI changed the route.[5]
CLOCKSPEED / ONE DOCUMENTED RECONFIGURATION
A change within an ongoing relationship
OpenAI's documented checkout sequence shows that a particular division of work can change within an ongoing relationship. Roughly five and a half months separated launch from reported withdrawal of the specified checkout route. The sequence challenges the assumption that an initial checkout position will persist. Investment in a particular integration becomes exposed when the channel changes the division of work before that investment has paid back.
EVENT 1 / 3 · 29 September 2025
Instant Checkout launch
The in-chat checkout route launched on 29 September 2025.
EVENT 2 / 3 · 9–12 March 2026
Withdrawal reported
Reports describe the withdrawal of catalog-merchant checkout, quoting OpenAI. Checkout components and embedded apps have separate configuration histories.
EVENT 3 / 3 · 24 March 2026
Shopify store route documented
Shopify describes ChatGPT discovery leading to checkout in the merchant’s store. Discovery and transaction infrastructure remain connected through a different route.
Which sources of advantage remain valuable when interfaces, operating capabilities and customer behavior change at different rates?
Which sources of advantage remain valuable across different partnership configurations?
Distinguish a position that depends on one channel design from a capability that several arrangements need. Merchant relationships, distinctive supply and reliable operating services are candidates for the latter; their durability still depends on substitutes and continued demand. Identify which contributions remain valuable as partners reorganize the customer journey, and which commitments depend on the current arrangement persisting.
Sources and configuration history
Held research: CP-2 v3, review packet E1 and Shopify deployed-to-deployed sequence; CP-9 v1.1, X-5 and C-3. Historical evidence cutoff: 7 October 2026. March withdrawal dates are report dates, not precise dates for every merchant. Instacart’s later continuity remains unresolved; the Shopify sequence must not be generalized to its app.
Inspect Shopify’s historical pilot and later referral route →Configuration boundaries
A new channel needs adoption and a complete service. The held research reports that OpenAI withdrew or reconfigured specified catalogue-merchant checkout routes in March 2026; termination is applied configuration by configuration, and Walmart’s in-ChatGPT checkout is separately bounded as ended no later than 30 March 2026. Google’s hosted checkout continued and was extended to hotels on 27 August 2026. Embedded provider apps and checkout components in ChatGPT are not shown to have ended with those routes. Buyers need sufficient supply and a working purchase journey before they will return.
EXPLORE THE RESEARCH
The arrangements behind the analysis
Inspect the recorded configuration and its competitive interpretation. Several records can describe the same partnership at different stages or in different roles.
Browse the full partnership research
Each record identifies the operating status and date of the arrangement.
CONNECT THE ARGUMENT
Strategic assessment
For a commerce platform, easier connections change the case for charging for integration alone. For a retailer, a new AI channel changes the value of customer access only as shoppers use it and return. Investment decisions therefore turn on whether the contribution remains useful when a partner changes the journey, and whether another route can preserve the result.
What to retain or build
Invest in differentiated supply, operating service and customer benefits—not only bespoke integration.
Which terms to accept
Protect account continuity, portability and the ability to use competing channels.
Other ways to explore Commerce
Research records
Based on the reconciled 81 episode/role records across 58 direct case bundles, including documented, planned, historical and specification material. Seven company-context cases and one boundary case remain outside the direct distribution. Source passages, dates and configuration history remain in the coordination-pattern record panels.
Sources and notes
Theory. Carliss Y. Baldwin and Marcus Holgersson, "Ecosystem Evolution: Contracts, Modules, and Standards," forthcoming handbook chapter, held manuscript, pp. 7-11, "Evolution via Modules" and "Evolution via Standards." Compatible modules and design rules can permit replacement. The Observatory applies this mechanism to connections with working compatibility and usable participation rights.
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, interaction models and interoperability. The discussion informs the distinction between reusable connections and underlying services. Its forecasts and a universal progression toward autonomous commerce are not adopted here.
Case evidence. Shopify, "Millions of merchants can sell in AI chats" (24 March 2026), merchant access, agentic storefronts and the merchant-store checkout route; Partnerships Observatory, Shopify standards and channel comparison, research cutoff 7 October 2026. The contribution comparison is the Observatory's interpretation.
Theory. Richard M. Emerson, "Power-Dependence Relations," American Sociological Review 27, no. 1 (1962), especially pp. 32-33. This motivates comparing the losses on each side and practical substitutes. The direction of bargaining change remains conditional on repeat demand, substitutability and usable alternatives.
Case evidence. OpenAI, "Buy it in ChatGPT" (29 September 2025); Shopify, "Millions of merchants can sell in AI chats" (24 March 2026); Partnerships Observatory, dated checkout reconstruction, research cutoff 7 October 2026. The reconstruction distinguishes launch, March withdrawal reports and the later Shopify store route. Report dates are not exact termination dates for every merchant; embedded apps have separate configuration histories.