Composable commerce with Medusa: start where the value is

We're part of the Medusa community and we build on it. But you don't always need a full replatform to get the benefits.

Composable commerce is less about choosing a new engine and more about modernising where the value is, not where the technology tells you to. That is part of our wider Experience Commerce thinking: we develop technology and architecture where they create the most value for the business and the customer experience.

It is gaining momentum in the Nordics again, and Medusa is a big part of the reason: an open-source, modular commerce core you can extend around your business instead of bending your business to fit a platform. We think the direction is right, and we have been moving with it.

We don't just use Medusa, we build it

Solteq is an active contributor to Medusa's core, its B2B capabilities and its commerce plugins. So when we build on Medusa, we are not learning it on your project. We help shape the platform itself, and that shows in how confidently we can extend it around real business needs.

Where Medusa fits

Medusa sits at the commerce core and connects to the systems you already run: ERP, PIM, CMS, search, CDP, OMS and fulfilment. It supports both B2C and B2B, multichannel and multilingual, and is Nordic-ready for payments and shipping. On the B2B side we build in organisation pricing, volume discounts and RFQ. And in the open-source setup there is no revenue-based platform fee, so cost stays stable as you grow instead of scaling with every sale.

It supports genuinely different business models: manufacturers selling through dealers and direct to consumers, multiple brands and markets, and demanding B2B. And because we build with a proven AI-assisted development model, standing these up and adapting them is fast.

You don't need a full replatform all at once

Most announcements describe a single full build on a new platform. Sometimes that is right. But often the commerce platform still works and there is no business case to replace it. In that case you can start with the storefront: a modern, independent storefront on top of your existing engine improves conversion, performance and accessibility now, and it decouples the customer experience from the platform's release cycle. No large migration required.

When the engine itself becomes the constraint, you move to a Medusa-based build from a foundation that is already proven. A single full build is one option. Often it is not the first one you need.

I wrote more about this frontend and composable shift in an earlier post: Composable's second coming: AI changed the equation

The proof, not just the promise

The storefront path is something we can also show, not just describe. In a recent proof of concept we built a server-rendered storefront on top of an existing commerce backend in under a month. In our own tests the front page scored 87 to 88 on mobile Lighthouse, with pages rendered per visitor on the server, ready for customer-specific pricing and personalization. That result is about the storefront and the way we build, not about any single engine, and we have gotten faster since.

Reactionary keeps your options open

We believe architecture should create options, not dependencies. That is why we built Reactionary, our open-source capability layer, which keeps both paths open. It sits between the storefront and the platforms, with providers behind a gateway, so integrations, caching and access control live in one place and the services behind them, Medusa included, can be swapped or added one at a time. And it is open source, not a proprietary accelerator you license.

The library already ships providers across the stack: Medusa, Commercetools, Adobe Commerce, Magento Open Source and HCL Commerce for commerce; Algolia and Meilisearch for search; Paytrail, Stripe and nShift Checkout for payments and delivery; Bazaarvoice and Lipscore for ratings and reviews; Google Analytics; and Apache Unomi for customer data. New providers are added as projects need them, and because they all follow one shared model, in practice any REST or GraphQL service can be connected.

COMPOSABLE COMMERCE

Reactionary: one foundation, many experiences, swappable backends

The same connected foundation also serves new agent channels, and that is where the competition is shifting. Reactionary exposes its capabilities to agents as APIs and supports the emerging agent protocols (MCP, ACP, UCP), so the store is both discoverable and transactable where buyers increasingly are: in AI assistants and shopping agents. An agent gets the same information a shopper has, with the same prices, assortment and business rules, and can search, compare, fill the cart and complete the purchase. Because this lives in Reactionary's shared layer rather than a single backend, it works the same across every engine we support, from one source of truth, with the storefront as just one channel among many.

What this means

We're active in the Medusa community and build on it, and we are glad the market is moving this way. But the engine is a choice anyone can make. The advantage is starting where the value is, proving it fast, and keeping what sits behind the experience open to change.

Composable commerce is not the destination. Business growth is. The objective is an architecture that lets you keep improving the customer experience, adopt new technology and pursue new revenue without rebuilding the foundation every few years. At Solteq we call this Experience Commerce. Medusa is one strong enabler on this journey.

If you are weighing Medusa, or composable commerce more broadly, that is the conversation worth having.

Want to see it in practice?

We'll show you our storefront approach, Reactionary and a live Medusa demo.

Get in touch →