Where the Money Goes Now: What Merchants Are Buying to Make Order Operations Ready for AI
In the first part of this piece we scored a decade of order management ideas and found a fairly consistent pattern: the ones that became operating practice changed how a retailer holds and moves data about stock, stores and orders, and the ones that stayed on the slide were labels for a platform purchase or changes to the customer-facing surface. This part is about the present. Retail technology budgets are growing in 2026, most of the growth is AI, and the question we wanted to answer with evidence rather than opinion is what, specifically, merchants are buying with it to make order operations ready for AI, and what they are quietly buying less of.
The money is going to AI, and most of it is not yet working
Start with the size of the shift. Gartner's July 2026 forecast put worldwide IT spending at $6.37 trillion in 2026, up 14.2%, with the growth attributed mainly to AI infrastructure. Its retail forecast (February 2026) put enterprise IT spending in retail at $232.8 billion in 2026, up 7.4%. In the supply chain, which is where order operations sit, Gartner's 2026 survey of 394 supply chain professionals found that 67% of supply chain digital investment now goes to AI, and a companion survey of 135 senior supply chain leaders found that 55% are unclear on the return (Gartner, August 2026). Among CIOs, Evanta's 2026 Leadership Perspectives survey of 1,400 IT executives found "Operationalizing AI" had overtaken cybersecurity, which had led for the previous four years, as the top functional priority.
Then look at what that money has produced so far. McKinsey's State of AI survey (November 2025) found that 88% of organisations use AI in at least one function, that nearly two-thirds have not begun scaling it across the enterprise, that 23% are scaling agents in at least one function, and that only 39% report any impact on earnings before interest and tax, most of it under 5%. Gartner said in June 2025 that over 40% of agentic AI projects will be cancelled by the end of 2027, and that of the thousands of vendors describing themselves as agentic, only about 130 were genuine. Deloitte's 2026 Global Retail Outlook, surveying 330 retail executives, found that 68% expect to deploy agentic AI for key operational work within 12 to 24 months, while TCS's Global Retail Outlook 2026, surveying more than 800 executives, found that 85% had not begun implementing multi-agent systems and that most effort was going into frontline chatbots.
So the budget has moved before the capability has. The useful question is why.
The reason is the data
The analysts are unusually unanimous on this. Gartner's February 2025 release "Lack of AI-Ready Data Puts AI Projects at Risk" predicted that "through 2026, organizations will abandon 60% of AI projects unsupported by AI-ready data", and reported from a survey of 1,203 data management leaders that 63% either did not have, or were unsure whether they had, the right data management practices for AI. Gartner research published in April 2026 found that organisations with successful AI initiatives invest up to four times more, as a share of revenue, in foundational capabilities such as data quality and data governance than organisations with poor AI outcomes (reported by BERI). Informatica's CDO Insights 2025 survey of 600 data leaders found 43% naming data as a top obstacle to moving generative AI from pilot to production, tied with technology. In retail specifically, Salesforce's 2025 Connected Shoppers Report (8,350 shoppers and 1,700 retail decision-makers) found 76% of retailers increasing their investment in agents, and noted in the same report that agents need unified data and that many retailers' systems remain disconnected. One widely shared recap of NRF's Big Show in January 2026 put it the same way: most retailers "are not truly ready to scale" agentic commerce, and "what's holding things back isn't imagination; it's data readiness" (Sitoo), while GS1 Sweden's recap made the narrower point that poor product data risks undermining agent-based commerce.
If that is the diagnosis, the spending should show up in the layer that holds the data, and it does.
What is being bought, layer by layer
Clean order and inventory data. The clearest statements come from the retailers with the largest store networks. Walmart's chief financial officer told CIO Dive that around 60% of US stores now receive freight from automated distribution centres and about 50% of eCommerce fulfilment centre volume is automated, which he said "enables better visibility into what inventory we own and inventory we can access", and chief executive John Furner described Walmart as "becoming AI native" on its first-quarter earnings call in May 2026 (Digital Commerce 360). Target attributed an improvement of more than 150 basis points in on-shelf availability for its 5,000 most important items, around 30% of unit sales, to machine learning in inventory planning in its fiscal third quarter of 2025 (Nasdaq), and in March 2026 announced a $2 billion plan that adds $1 billion of capital spending and $1 billion of operating expense to improve the customer experience and speed up technology adoption (CIO Dive). Forrester's May 2025 note on the OMS Wave put the industry version of the same point plainly: an OMS that does not provide near-real-time inventory availability "just won't cut it anymore".
Event streams and APIs. Confluent's 2025 Data Streaming Report, a vendor survey of 4,175 IT leaders in 12 countries (May 2025), found that 90% planned to increase spending on data streaming platforms and that 89% said those platforms eased AI adoption by addressing data access, quality and governance. The pattern of doing the plumbing before the AI is visible in individual companies. Wayfair's chief executive told investors on its third-quarter 2025 call (28 October 2025) that "our multiyear replatforming is largely complete" after "about three years on a very large technology replatforming effort for almost all of our core systems", and its chief technology officer said in the same call that the plan was to make the catalogue "fully transactable on leading AI platforms". The order of those two statements is the point.
The orchestration layer. The order management vendors have spent 2026 putting agents inside the system rather than beside it. Manhattan Associates made an OMS Configuration Agent, a Store Associate Agent and a Contact Center Agent commercially available in Active Omni in January 2026, with Brian Kinsella, its SVP of product management, saying "we've long believed true AI needs to live within rather than alongside our applications". Blue Yonder introduced a Fulfillment Agent that explains sourcing decisions in its order management product the same month. IBM released a Sterling OMS Agentic Toolkit in March 2026, with agents for order information, cancellations and inventory segmentation. Salesforce announced Agentforce Commerce with "Agentic Order Management" for routing, fulfilment and disruption alerts in June 2026, and Kibo made KIBO AI generally available in August 2026 on a single data model across commerce and order management. Nucleus Research's 2026 OMS Value Matrix summed up the shift: agentic AI "has moved from roadmap to production across the market", and the OMS is "the connective orchestration layer" between storefronts, stores, carriers and financial systems "rather than a back-office order book".
Agent access to order systems. This is the newest line in the budget and the one with the most public standards behind it. Anthropic released the Model Context Protocol in November 2024; OpenAI adopted it in March 2025; by December 2025 the Linux Foundation's new Agentic AI Foundation reported more than 10,000 published MCP servers, with Shopify, Salesforce, SAP, Snowflake and Adyen among its members. Order management vendors followed: Pipe17 shipped an MCP server for order operations in September 2025, commercetools' Commerce MCP exposes tools to read, create and update orders, and Kibo launched KIBO MCP in January 2026. Shopify's developer documentation now describes an Order MCP for order tracking alongside its cart and checkout endpoints. Two things in the commerce protocols are worth noticing. The Stripe and OpenAI Agentic Commerce Protocol announcement of September 2025 says merchants remain the merchant of record and "handle fulfillment and returns, as they normally would". And when Google and Shopify launched the Universal Commerce Protocol at NRF in January 2026, with Target, Walmart, Wayfair and Etsy among the launch partners, Wayfair's chief technology officer described it as "the common language for this new ecosystem" while Wayfair "remains the merchant of record". The agent handles discovery and checkout. The merchant keeps the order, the stock, the delivery and the return, and returns are not small: Salesforce put online returns over the 2025 holiday season at $181 billion, 14% of purchases, in its vendor-reported recap.
The volumes through these channels are still modest, and the companies involved say so. Shopify's president told investors in November 2025 that AI-driven traffic to Shopify stores was up seven times since January and orders attributed to AI searches up eleven times; by February 2026 the order multiple was fifteen, "on a small base", and The Logic reported that agentic commerce had yet to make a meaningful impact on Shopify's results (PYMNTS; The Logic). Adobe reported generative-AI-referred traffic to US retail sites up 693.4% year on year over the 2025 holiday season, from a small base, and Salesforce attributed $262 billion, 20% of global online holiday sales, to AI and agents, a vendor figure that counts any AI-influenced interaction. The incumbents are also gating the door: Amazon sued Perplexity in November 2025 over its shopping agent's access to customer accounts. The access layer is being built, but it is being built behind permissions, which is itself a statement about where the valuable data sits.
What is being bought less
The quieter half of the story is the storefront. Commerce.com, the renamed BigCommerce, told investors on its second-quarter 2026 call that "B2C replatforming activity remains softer than we've seen historically", that AI "is changing how merchants evaluate technology investments and delaying monetization across portions of the industry", and that deals were taking longer as merchants reassessed priorities around AI-driven discovery; it cut its revenue outlook and shifted investment toward AI, product intelligence and payments (Nasdaq). Forrester's 2026 digital commerce predictions (28 October 2025) expect one-third of retail marketplace projects to be deserted as answer engines take the traffic, with the freed spend reallocated to earning positions in those engines. The MACH Alliance's own 2023 research recorded the share of respondents prioritising front-end investment falling from 54% in 2022 to 39% in 2023. And Shopify, which has more reason than anyone to sell storefronts, published a post in April 2026 on the "integration tax" of fragmented stacks whose central line was that "AI won't solve fragmented architecture. It exposes it."
None of these is a survey that says storefront budgets were cut, and we have not found one. What they show, together, is that the discretionary decision to rebuild the customer-facing surface is being deferred while the discretionary decision to fix the data underneath it is being funded.
What the evidence adds up to
Put the pieces in order. Budgets are up and the increase is AI. Most AI projects are stalling, and the analysts who track them say the cause is data that is not ready. The retailers with the best operating results are spending on inventory accuracy, automated fulfilment and the systems that feed them. The order management vendors are rebuilding their products as orchestration layers with agents inside. The protocols that let outside agents sell on a merchant's behalf explicitly leave the order, the fulfilment and the return with the merchant. And the one category whose demand is visibly softening is the storefront re-platform.
We do not think this is a prediction. It is a description of where the money already went in 2025 and 2026, and it says that for AI in commerce the storefront is becoming one channel among several, some of which the merchant will not own, while the thing the merchant does own, and that an agent in any channel will need to ask about, is the order: where the stock is, where the order will ship from, when it will arrive, and what happens if it comes back. That is where the investment is going, and on the evidence of Part 1 it is also the kind of investment that has tended to last.
For what it is worth, this is the layer we work in at Carriyo, so readers should weigh our reading of the evidence with that in mind. The sources are below and the argument should stand or fall on them.
---
Sources
1. Gartner. "Gartner Forecasts Worldwide IT Spending to Grow 14.2% in 2026, Totaling $6.37 Trillion", 27 July 2026. https://www.gartner.com/en/newsroom/press-releases/2026-07-27-gartner-forecasts-worldwide-it-spending-to-grow-14-point-2-percent-in-2026-totaling-6-point-37-trillion 2. Gartner. "Forecast: Enterprise IT Spending for the Retail Market, Worldwide, 2023-2029, 4Q25 Update", 4 February 2026 (abstract). https://www.gartner.com/en/documents/7401230 3. Gartner. "Gartner Survey Finds Majority of Chief Supply Chain Officers Unclear on AI Investment Returns", 5 August 2026; republished by Supply Chain 24/7. https://www.gartner.com/en/newsroom/press-releases/2026-08-05-gartner-survey-finds-majority-of-chief-supply-chain-officers-unclear-on-ai-investment-returns ; https://www.supplychain247.com/article/gartner_cscos_unclear_on_ai_roi/gartner 4. Evanta. "Top 3 priorities for CIOs in 2026", Leadership Perspective Survey of 1,400 IT executives (May 2026). https://www.evanta.com/resources/cio/survey-report/top-3-priorities-for-cios-in-2026 5. McKinsey & Company. "The state of AI in 2025: Agents, innovation, and transformation", November 2025; figures as reported by IT Brief, 11 November 2025. https://www.mckinsey.com/capabilities/quantumblack/our-insights/the-state-of-ai ; https://itbrief.co.uk/story/mckinsey-report-shows-ai-interest-but-slow-scaling 6. Gartner. "Gartner Predicts Over 40% of Agentic AI Projects Will Be Canceled by End of 2027", 25 June 2025, as reported by New Electronics. https://www.newelectronics.co.uk/content/news/over-40-of-agentic-ai-projects-could-be-cancelled-by-2027 7. Deloitte. "2026 Global Retail Outlook". https://www.deloitte.com/content/dam/assets-shared/docs/industries/consumer/2026/global-retail-outlook-2026.pdf 8. TCS. Global Retail Outlook 2026, as reported by Yahoo Finance. https://finance.yahoo.com/news/retailers-worldwide-lag-adopting-advanced-013921808.html 9. Gartner. "Lack of AI-Ready Data Puts AI Projects at Risk", 26 February 2025. https://www.gartner.com/en/newsroom/press-releases/2025-02-26-lack-of-ai-ready-data-puts-ai-projects-at-risk 10. BERI. "Gartner: AI success linked to 4x investment in data and analytics foundations", 18 April 2026 (reporting Gartner research of 16 April 2026). https://www.beri.net/article/gartner-ai-success-4x-data-analytics-investment-2026 11. Informatica. "CDO Insights 2025", survey of 600 data leaders. https://www.informatica.com/blogs/cdo-insights-2025-global-data-leaders-racing-ahead-despite-headwinds-to-being-ai-ready-latest-survey-finds.html 12. Salesforce. "Connected Shoppers Report", sixth edition, 2025. https://www.salesforce.com/news/stories/ai-agent-retail-trends-2025/ 13. Sitoo. "Top trends and takeaways from NRF 2026"; GS1 Sweden, NRF 2026 recap. https://sitoo.com/resources/top-trends-and-takeaways-from-nrf-2026 ; https://gs1.se/en/?p=71624 14. CIO Dive. "Walmart's AI wins" (CFO John David Rainey on automated distribution and inventory visibility). https://www.ciodive.com/news/walmart-AI-wins-sparky-john-furner/812606/ 15. Digital Commerce 360. "Walmart's Sparky agent, AI sales and supply chain", 22 May 2026. https://www.digitalcommerce360.com/2026/05/22/walmart-sparky-agent-ai-sales-supply-chain/ 16. Nasdaq. "AI becoming the backbone of Target's ambitious retail turnaround" (on-shelf availability, Q3 2025). https://www.nasdaq.com/articles/ai-becoming-backbone-targets-ambitious-retail-turnaround 17. CIO Dive. "Target's $2B investment in AI and technology" (Q4 2025 call, 4 March 2026). https://www.ciodive.com/news/target-AI-investments-michael-fiddelke/813839/ 18. Pfeiffer, E. "Highlight Reel: Insights From The Latest OMS Forrester Wave", Forrester, 8 May 2025. https://www.forrester.com/blogs/highlight-reel-insights-from-the-latest-oms-forrester-wave/ 19. Confluent. "2025 Data Streaming Report", 20 May 2025. https://www.confluent.io/press-release/data-streaming-report-2025/ 20. Wayfair. Third quarter 2025 earnings call transcript, 28 October 2025. https://allmind.ai/earnings/event/2815364 21. Business Wire. "Manhattan Associates announces commercial availability of agentic AI in Manhattan Active Omni", 9 January 2026. https://www.businesswire.com/news/home/20260109780453/en 22. Blue Yonder. "On the road to cognitive, intelligent and agentic Blue Yonder Order Management", 9 January 2026. https://blueyonder.com/blog/2026/on-the-road-to-cognitive-intelligent-and-agentic-blue-yonder-order-management 23. IBM. "Introducing agentic AI for order management with IBM Sterling OMS", March 2026. https://www.ibm.com/new/announcements/introducing-agentic-ai-for-order-management-with-ibm-sterling-oms 24. Salesforce. "Agentforce Commerce" announcement, 24 June 2026. https://www.salesforce.com/news/stories/agentforce-commerce-announcement/ 25. Kibo. "KIBO AI generally available", August 2026. https://kibocommerce.com/press-events/kibo-ai-unified-agentic-commerce-oms/ 26. Nucleus Research. "OMS Technology Value Matrix 2026", 21 April 2026. https://nucleusresearch.com/research/single/oms-technology-value-matrix-2026/ 27. TechCrunch. "OpenAI adopts rival Anthropic's standard for connecting AI models to data", 26 March 2025. https://techcrunch.com/2025/03/26/openai-adopts-rival-anthropics-standard-for-connecting-ai-models-to-data/ 28. Linux Foundation. "Linux Foundation Announces the Formation of the Agentic AI Foundation", 9 December 2025. https://www.linuxfoundation.org/press/linux-foundation-announces-the-formation-of-the-agentic-ai-foundation 29. Pipe17. "2025 product year in review" (MCP server, September 2025). https://pipe17.com/blog/2025-product-year-in-review/ 30. commercetools. Commerce MCP documentation. https://docs.commercetools.com/sdk/mcp/commerce-mcp 31. Kibo. "Kibo launches Connect Hub and MCP", 12 January 2026. https://kibocommerce.com/press-events/kibo-launches-connect-hub-and-mcp-2026/ 32. Shopify. Storefront MCP developer documentation (Order MCP). https://shopify.dev/docs/apps/build/storefront-mcp 33. Stripe. "Stripe and OpenAI launch Instant Checkout and the Agentic Commerce Protocol", 29 September 2025. https://stripe.com/newsroom/news/stripe-openai-instant-checkout 34. Digital Commerce 360. "Wayfair on Google's Universal Commerce Protocol", 14 January 2026; The New Stack, "Shopify, Walmart endorse Google's new open commerce protocol". https://www.digitalcommerce360.com/2026/01/14/wayfair-google-universal-commerce-protocol/ ; https://thenewstack.io/shopify-walmart-endorse-googles-new-open-commerce-protocol/ ; https://ucp.dev 35. Salesforce. "Holiday 2025 shopping data", 8 January 2026. https://www.salesforce.com/news/stories/2025-holiday-shopping-data/ 36. PYMNTS. "Shopify president: agentic AI drives 11x spike in orders", 4 November 2025; The Logic, Shopify Q4 2025 earnings, February 2026. https://www.pymnts.com/news/ecommerce/2025/shopify-president-agentic-ai-drives-11x-spike-in-orders ; https://thelogic.co/news/shopify-earnings-q4-2025-artificial-intelligence/ 37. Adobe. "Adobe: 2025 holiday shopping season", 7 January 2026. https://news.adobe.com/news/2026/01/adobe-holiday-shopping-season 38. Retail Dive. "Amazon sues Perplexity over AI shopping agent", November 2025. https://www.retaildive.com/news/amazon-sues-perplexity-ai-shopping-agents/804871/ 39. Nasdaq. "CMRC Q2 earnings call flags softer B2C demand, outlook cut", 2026. https://www.nasdaq.com/articles/cmrc-q2-earnings-call-flags-softer-b2c-demand-outlook-cut 40. Pfeiffer, E. "Predictions 2026: The 'Agentic Commerce' Race, And Some Potential Regrets, In Digital Commerce", Forrester, 28 October 2025. https://www.forrester.com/blogs/predictions-2026-the-agentic-commerce-race-and-some-potential-regrets-in-digital-commerce/ 41. MACH Alliance. "MACH Global Research 2023", 22 March 2023. https://machalliance.org/newsroom/mach-global-research-2023 42. Shopify. "The frankenstack: the hidden cost of fragmented commerce architecture", 24 April 2026. https://www.shopify.com/enterprise/blog/frankenstack