Commerce is moving toward AI-led buying, where AI influences what customers discover, compare, and purchase. McKinsey estimates that agentic commerce could orchestrate $3 trillion to $5 trillion in global consumer commerce by 2030, while 44% of people who have tried AI-powered search, prefer it over than traditional search.
These numbers highlight an important issue for enterprise leaders. Are their commerce systems ready for AI to play a bigger role in product discovery and buying decisions? This means adding AI to an existing storefront is not enough. Enterprises need to connect their commerce systems, data, and operations so AI can access the information it needs to make reliable decisions.
Changes that will reshape Enterprise Commerce
-
Commerce will move from channels to capabilities
Customers may discover products through an AI assistant, interact with a conversational commerce application, purchase through a marketplace, or work with a sales representative using an AI-powered configuration tool. Every interaction depends on accurate product information, pricing, inventory availability, customer eligibility, ordering, payment, fulfillment, and returns.
This creates a shift from channel-centric commerce to capability-centric commerce. Instead of building commerce logic around individual interfaces, enterprises need reusable capabilities that can be used across channels and applications.
-
AI agents will increasingly participate in commerce decisions
The role of AI in commerce will go beyond generating product descriptions, powering search, or recommending products. As AI capabilities mature, these AI agents will increasingly help guide customers from intent to purchase.AI may help customers discover and compare products, then provide recommendations based on individual requirements, and execute selected actions on customer’s behalf.
-
Commerce decisions will depend on operational data
Consider a customer searching for a product. To provide an accurate recommendation, system needs to understand customer’s account, current pricing, inventory position, expected demand, fulfillment capacity, and delivery options.
How to prepare your commerce stack for AI?
If inventory information is outdated, the recommendation may be inaccurate. As a result, systems such as PIM, CRM, ERP, OMS, inventory, pricing, and fulfillment platforms influence customer experience.
-
Make product data ready for machine-driven commerce
Product data must support both human buyers and AI systems. AI needs structured information about product attributes, relationships, compatibility, pricing, availability, delivery, and policies. In B2B commerce, this includes account-specific catalogs, negotiated pricing, technical specifications, and customer eligibility. Enterprises should treat product data as a commerce decision asset.
-
Expose commerce capabilities independently of the storefront
Enterprises should avoid rebuilding commerce functionality for every new channel. Capabilities such as product discovery, pricing, inventory, customer identity, checkout, orders, and returns should be available as reusable services across websites, marketplaces, mobile apps, sales tools, and AI interfaces. API-first and composable commerce provide flexibility to extend these capabilities as commerce models evolve.
-
Connect customer experience with fulfillment
Connecting PIM, commerce, inventory, OMS, and fulfillment systems ensures that product recommendations, availability, pricing, and delivery promises are based on current information. As AI takes on more commerce decisions, reliable data becomes essential to prevent inaccurate recommendations and unfulfilled customer promises.
-
Establish clear boundaries for AI autonomy
Enterprises should define which decisions AI can execute, which require human oversight, and which need approval. Low-risk recommendations may be automated, while pricing changes, high-value transactions, refunds, and other sensitive decisions may require human control. AI autonomy should be determined by business risk and impact.
-
Rethink how commerce performance is measured
Conversion, revenue, and average order value will remain important, but enterprises also need to measure the impact of AI and automation. Relevant metrics include AI-assisted revenue, automated decision rates, cost per transaction, order exception rates, fulfillment accuracy, margin impact, and time to decision.
Conclusion
Commerce leaders do not need to predict how customers will buy in 2030. They need to ensure their businesses can respond when those behaviors, channels, and technologies change.
This implies moving beyond isolated storefronts and point solutions toward a connected commerce ecosystem in which product data, customer intelligence, pricing, inventory, order management, and fulfillment work together. It also means creating the flexibility to introduce AI and new commerce experiences without replacing the underlying technology stack.
For enterprises, this is where Ignitiv’s approach to commerce modernization becomes relevant. We focus on connecting systems, data, and capabilities that already power the business and creating a foundation for innovation.
FAQs
AI agents will work beyond product recommendations to evaluate products, pricing, availability, customer-specific rules, and delivery options. This will require enterprises to expose reliable commerce data and capabilities to machine-driven interfaces.
Adding every new channel as a separate implementation increases integration complexity and creates inconsistent customer experiences. Reusable commerce capabilities allow websites, marketplaces, sales portals, and AI interfaces to use the same products, pricing, inventory, and order services.
The OMS can provide critical information about order status, inventory allocation, fulfillment, and exceptions. This allows AI-driven experiences to make recommendations and promises based on what the enterprise can fulfill rather than relying only on storefront information.
AI-ready product data needs more than descriptions and images. It should include structured attributes, product relationships, compatibility, pricing, availability, promotions, delivery information, and policies in formats that systems can interpret.
Start with simple, repetitive, low-risk tasks such as product recommendations, search rankings, customer service responses, and merchandising suggestions. Keep higher-risk decisions, such as pricing changes, large B2B orders, refunds, and credit decisions, under human oversight.
