Artificial intelligence is rapidly becoming embedded in the retail technology stack. What started as experimentation is now moving into operational adoption as retailers look for ways to improve efficiency across digital commerce teams.  

According to Nvidia’s 2026 State of AI in Retail report, 91% of retailers are already using or evaluating AI and 90% plan to increase investment in the technology, with most organisations reporting improvements in both revenue and cost reduction.

Yet scaling AI across commerce operations remains difficult. Research shows that more than seventy percent of retailers are piloting AI agents while only eight percent have fully deployed them across their operations.  

The gap between experimentation and operational impact often comes down to one challenge: connecting AI systems to the complex technology environments that power modern commerce.

‍

Why AI struggles inside retail technology ecosystems

Retail technology ecosystems are built on multiple platforms. Commerce engines such as Salesforce Commerce Cloud sit alongside product information management systems, order management platforms, marketing technologies and internal data environments. These systems contain the operational data that powers merchandising, customer journeys, product content and fulfilment.

For AI to deliver meaningful productivity gains it must be able to interact with these systems in real time. However, most enterprise platforms were not designed to work directly with AI agents. As organisations attempt to scale AI across their digital operations, they quickly encounter friction, particularly when models need to access live operational data across different platforms.

‍

The integration problem slowing AI innovation

When organisations introduce AI tools, engineering teams often need to build custom integrations so those systems can access operational data. Product catalogues, customer information and order data typically sit across different platforms, and each AI model may require separate connections to retrieve or interact with that information.

Technology analysts often describe this as the “N by M integration problem,” where every AI system needs separate integrations with each enterprise platform. As organisations experiment with more AI tools the number of integrations grows quickly, increasing technical complexity and slowing development cycles across digital teams.

‍

The emerging role of Model Context Protocol

For AI to become operational inside commerce teams it needs a structured way to interact with enterprise systems. The Model Context Protocol (MCP) is an open standard designed to allow AI systems to securely connect with external tools, applications and data sources through a consistent interface.

Instead of building new integrations every time an AI system needs to access a platform, organisations can expose system capabilities through MCP servers. These servers act as a controlled interface between AI agents and enterprise systems, allowing models to retrieve data, interact with APIs and support operational workflows using live business information.

‍

Enabling AI within Salesforce Commerce Cloud

For organisations running Salesforce Commerce Cloud, this shift opens new opportunities to connect AI capabilities directly with commerce operations. Commerce platforms sit at the centre of digital retail ecosystems, managing product catalogues, customer journeys and transactional data. Enabling AI to interact with this environment allows organisations to explore new productivity improvements across both engineering and commerce teams.

At The Commerce Team Global, we developed a Salesforce B2C MCP Server designed to expose Salesforce Commerce Cloud capabilities through the Model Context Protocol. This provides a structured integration layer between AI systems and commerce platforms, helping organisations experiment with AI driven workflows without introducing repeated custom integrations.

You can learn more about the solution here.

‍

Moving from AI experimentation to operational productivity

The conversation around AI in retail is evolving quickly. Early use cases focused on experimentation and content generation, but the next phase will focus on operational productivity. Retailers are increasingly exploring how AI can interact directly with the systems that run digital commerce.

Protocols such as MCP provide a pathway to connect AI with enterprise platforms in a scalable way. For organisations looking to move beyond experimentation, establishing the right integration architecture will be critical to unlocking the productivity benefits that AI promises across modern commerce operations.

‍

How TCTG Can Help

Retailers exploring AI are increasingly discovering that the real challenge is not the model itself, but how AI connects with the systems that power commerce operations. Establishing the right architecture early makes the difference between isolated pilots and scalable productivity.

If you want to understand how MCP can enable AI within Salesforce Commerce Cloud environments, we would be happy to share our experience working with commerce teams facing this challenge.

Contact us at info@thecommerceteam.com or connect with us on LinkedIn.

‍

‍