Digital Insights: Agentic Commerce

The digital retail world is on the brink of radical change. While websites, online stores and traditional customer journeys still dominate today, there are many indications that we are at the beginning of a profound paradigm shift. Agentic commerce is changing the way customers shop and therefore digital commerce itself.
AI agents and large language models (LLMs) are increasingly taking over entire purchasing processes - from determining requirements to researching, selecting products and placing orders. Customers no longer necessarily interact with a store front end, but with an AI agent that autonomously makes decisions, compares prices and executes orders.

From the storefront to the AI agent: A vision of the future

We are already seeing a significant decline in traditional website visits. Users are increasingly asking questions directly to LLMs instead of search engines. These models consume, filter and evaluate information independently, changing the foundations of digital commerce. In a few years' time, most efficiency-driven purchases could be made via AI agents, doing away with store front-ends altogether.
While browsing, inspiration and brand experiences will remain particularly relevant in B2C, efficiency will be the dominant criterion in many areas, such as

• B2B procurement
• MRO/ C-Parts and fast-moving products
• Regular consumables
• Products with clear technical parameters

The decisive factor here is not an intensive shopping experience, but speed, security and precision. And this is where AI agents are unbeatable.

Why modern system architectures are now making the difference

To enable AI agents to address systems directly, companies need technological prerequisites that only modern architectures can offer:

• API-first commerce
• Headless & composable systems
• microservices
• MCP interfaces (Model Context Protocol)
• Real-time data provision

These architectural principles form the technical basis so that agents can not only read data but also carry out transactions.
Technology partners such as Shopware, commercetools and Shopify are already taking up this development.

Agentic commerce is not a store issue - it is a data issue

The real disruption caused by agentic commerce is not taking place in the front end, but in data management. LLMs can only decide what they understand. And they only understand what has been recorded in a structured, complete and semantically clean way. This puts PIM systems (Product Information Management) and master data architectures at the center of the next wave of innovation. Companies must:

• Standardize product attributes
• Clearly record technical features
• Provide contextual information
• standardize taxonomies
• Maintain compatibilities, standards and application scenarios

This is the only way AI agents can reliably compare, evaluate and recommend products.

LLMO: Visibility in the agent ecosystem

With Agentic Commerce, a new discipline is gaining in importance: LLMO - Large Language Model Optimization. While SEO optimizes websites for search engines, LLMO ensures that AI agents can understand, compare and select products. This means that products, attributes, categories and technical data must be structured in such a way that an LLM can interpret them unambiguously. LLMO is the key to being present in agent decisions and the logical evolution of SEO.

Conclusion: Change is coming.

Agentic commerce is changing the entire value chain of digital commerce:
Data, APIs, PIM, integration, architecture and customer expectations. Companies that act early will secure the best places in a market in which AI agents, rather than people, will select products. We support you in preparing for this change in the best possible way.

Christof Zahn, Director Business Unit

Wir freuen uns, Sie kennenzulernen.