Amazon Business hits $60 billion in annualized sales as AI and computer vision reshape B2B ecommerce in 2026

Amazon Business has reached $60 billion in annualized gross sales, according to Digital Commerce 360, cementing its position as a dominant force in B2B digital commerce and raising the competitive bar for every procurement platform, distributor portal, and supplier marketplace operating today. That figure lands in the same week Instacart announced it has acquired Arpalus, a computer vision company focused on grocery shelf intelligence, and as Deloitte guidance reported by Digital Commerce 360 urges B2B ecommerce operators to restructure their catalog data before agentic AI agents make those decisions for them.

Amazon Business at $60 billion: what the number means for B2B operators

The $60 billion annualized gross sales milestone for Amazon Business, reported by Digital Commerce 360's Abbas Haleem, reflects years of steady expansion into business purchasing categories ranging from office supplies and MRO to lab equipment and industrial goods. For procurement leaders, the scale matters less as a headline and more as a competitive signal: suppliers that have not built out Amazon Business storefronts, configured punch-out catalogs, or enabled business pricing tiers are increasingly invisible to a buyer base that has normalized the platform.

Forbes has tracked the broader ecommerce market context, noting that global ecommerce continues to capture a growing share of total retail and B2B sales. Amazon's B2B channel has grown faster than its consumer marketplace in recent periods, reflecting a structural shift in how corporate buyers source goods. For operations and procurement leaders, the practical implication is straightforward: the catalog, the pricing, and the fulfillment SLAs a company offers on Amazon Business increasingly define its market reach among business buyers.

The $60 billion figure also reinforces what procurement teams managing vendor portfolios already sense: consolidation is happening at the buyer interface level. Fewer portals, more volume through fewer platforms, and greater pressure on suppliers to meet platform-specific data and fulfillment standards.

Instacart acquires Arpalus to bring computer vision into grocery operations

Instacart's acquisition of Arpalus, reported by Digital Commerce 360's Mary Meisenzahl on July 21, brings computer vision technology directly into the grocery fulfillment and retail intelligence stack. Arpalus specializes in shelf-level product recognition, using image analysis to track product placement, identify out-of-stocks, and validate planogram compliance in real time. For grocery retail operators that partner with Instacart for same-day delivery and in-store picking, the integration could meaningfully reduce the gap between what a retailer's inventory system says is on the shelf and what a shopper picker actually finds.

Out-of-stock substitutions are one of the most persistent sources of friction in grocery ecommerce fulfillment. When a picker cannot find an item, the customer experience degrades and the retailer absorbs the cost of substitution or cancellation. Computer vision at the shelf level addresses the problem upstream, before a pick order is dispatched, by giving operators a more accurate real-time picture of shelf state. Whether Instacart integrates Arpalus capabilities into its Storefront Pro platform for retail partners or deploys them in its fulfillment center operations remains to be detailed, but the acquisition signals a clear direction: physical retail shelf data will become part of the digital fulfillment loop.

For grocery retail operations directors evaluating their fulfillment partnerships, the Arpalus acquisition adds a layer of shelf-intelligence capability to the Instacart platform that few standalone retailers could build internally. It also raises a longer-term question about data ownership: as computer vision systems capture granular shelf and product data, retailers will need to understand how that information is used, retained, and potentially shared across the platform's network.

Agentic AI is changing how B2B buyers find products, and most catalogs are not ready

The third major development reported by Digital Commerce 360 this week, drawing on Deloitte analysis, focuses on agentic AI and its effect on B2B product discovery. Agentic AI refers to AI systems that can autonomously execute multi-step tasks on behalf of a user, including searching for, evaluating, and in some cases purchasing products without direct human input at each step. As these systems move from pilot to production inside enterprise procurement workflows, the way a product is described, attributed, and structured in a seller's catalog determines whether an AI agent can find and recommend it at all.

The operational gap is significant. Many B2B ecommerce catalogs were built for human keyword search, with product titles, descriptions, and attributes that rely on a buyer's intuition to navigate. Agentic AI agents parse structured data differently. They depend on consistent taxonomy, rich attribute fields, and machine-readable specifications to match a buyer's stated requirement to an available product. A catalog that works adequately for a trained procurement professional searching manually may be nearly invisible to an AI agent executing an autonomous sourcing run.

Digital Commerce 360's coverage of the Deloitte guidance emphasizes that B2B ecommerce teams should audit their product data architecture now, before agentic purchasing becomes the norm rather than the exception. That means reviewing attribute completeness, normalizing units and terminology across product families, and ensuring that structured data formats align with the standards that AI purchasing systems are being trained to read. Companies that delay this work risk being filtered out of AI-generated shortlists before a human buyer ever enters the process.

What ecommerce leaders should watch across all three developments

These three developments, Amazon Business's scale milestone, Instacart's computer vision acquisition, and the agentic AI catalog imperative, are distinct stories but share a common thread. Digital channels are becoming the primary operating surface for commerce, and the quality of data, the richness of platform integrations, and the accuracy of real-time inventory are now direct drivers of operational performance, not just marketing metrics.

Forbes's ecommerce statistics research, audited as of July 2026, provides additional context on the scale of the shift: ecommerce's share of global retail continues to climb, reinforcing that the infrastructure decisions operators make today, around catalog structure, fulfillment partnerships, and AI readiness, compound in significance as more purchasing volume moves online. For a VP of Operations or CIO evaluating technology investment priorities this year, the message from this week's news is consistent: the margin for delay on digital commerce infrastructure is shrinking.

Instacart has not disclosed financial terms for the Arpalus acquisition. Amazon Business has not separately broken out its B2B segment growth rate beyond the gross sales figure. Deloitte's agentic AI readiness recommendations, as reported by Digital Commerce 360, are expected to be elaborated in forthcoming guidance for enterprise ecommerce teams.