Commerce 2.0
The world's most flexible commerce platform, open-source and the evolution of software into internet memes.
How Big is E-Commerce?
ChatGPT values it at $6.3 trillion, Claude forecasts $8.1 trillion by 2026, and Perplexity even suggests a sky-high $16 trillion+ (or maybe just hallucinates more). Whatever the exact number, e-commerce is massive and still growing fast, now making up about 20% of total retail sales. At our AGM yesterday, my partner Robin reminded LPs that at the start of the internet, even the wildest estimates of e-commerce’s potential were much lower than these numbers. So, what’s next?
In the first ascent of e-commerce, the goal was simply to be online. Having an online store quickly transitioned from a competitive advantage to becoming an existential necessity for retailers. At the same time, it created an unprecedented opportunity for a new generation of entrepreneurs, providing them with a global distribution channel for any product they could imagine—removing the constraints of physical locations. Alongside Amazon’s marketplace, Shopify became the key enabler and beneficiary of this wave (in the Western world).
Through a combination of a “business-in-a-box” product, powerful network effects (an app store of partners and then the Shop app), and relentless execution, Shopify built a $140 billion+ juggernaut. It’s an amazing company that I admire greatly. However, I believe we’re now entering a new era for e-commerce, and while Shopify will remain a large player, I don’t think their approach is best suited to capture the same share of what’s coming.
As I mentioned, the first e-commerce era was about going online. Today, about 20% of total retail sales happen online, but we’re seeing a shift as a few converging factors reshape the landscape while expanding possibilities:
Retail is more competitive than ever: In an increasingly noisy market, brands and retailers must stand out through bespoke customer experiences. This requires a commerce platform capable of creating unique and tailored interactions. The differentiation potential is growing rapidly with the rise of spatial computing, 3D interfaces, and other evolving digital and physical form factors.
Commerce is becoming more holistic: we’re transitioning from e-commerce (a digital distribution channel) to digital commerce, where the entire business operates digitally. Whether transactions and fulfilment happen online, in-store, or in a hybrid format, the processes must be digitised and seamlessly integrated. This requires commerce platforms that adapt to business workflows rather than forcing businesses to adapt to the standard platform.
B2B commerce is huge but lagging: B2B online commerce market is estimated at approx $30 trillion with an expected CAGR of 17% for the next 5 years. However, the proportion of total B2B online transactions remains low compared to the B2C sector. B2B transactions are often more complex and require bespoke workflows that standard SaaS platforms couldn’t handle. As a result, many B2B platforms are still built from scratch or dominated by service industries.
The new tech stack and AI wave: technology is becoming modular, like Lego. APIs, distributed systems, and now AI are lowering the barriers to software creation while expanding its capabilities. This presents retailers, entrepreneurs, and businesses with new opportunities to rethink their digital commerce strategies and architectures.
All these factors not only create a massive opportunity for new software products but also directly challenge the dominant SaaS business model, which was built around generalisation and standardisation. The shift is now inevitably moving in the opposite direction: customisation and flexibility.
If you agree with these observations and premise, you’ll probably share my excitement for our portfolio company Medusa and the launch of their 2.0 release last month—a 12-month+ labor of love. With this release, Medusa has created the world’s most flexible commerce platform. They’ve fully modularized the core components (17 modules including cart, pricing, fulfilment, etc.), allowing anyone to customise the commerce experience to their needs. Beyond the modularity they’ve enabled, they created an entirely new framework. Think of it as Lego for commerce: take the essential building blocks, assemble them according to your logic, and integrate them seamlessly with any software you already use. They let you build any commerce feature to fit any use case.
Already the most popular commerce platform on GitHub, Medusa 2.0 enables cutting-edge commerce experiences for consumer and B2B use cases. Here are some examples:
Heineken in APAC: Powers a distributor platform handling over 10,000 daily orders.
Marvel Comics Store: Handles 1 million cart requests per minute (all with sub-500ms latency and 0% cart failures). User story coming soon.
Mitsubishi: Personalizes the online experience for new car purchases.
Makro Pro: Manages omnichannel order orchestration with over 5,000 daily orders.
Viessman: the global leader in heating, ventilation and air conditioning, launched a marketplace for mobile heating and cooling rentals.
Catalog: (another LG portfolio company) built a platform to digitise the b2b ordering process for industrial SMBs.
The greatest tech companies expand the realm of possibilities by creating breakthrough products. As we can see from the examples above, Medusa is well-positioned to power new use cases and redefine digital commerce.
But beyond commerce, Medusa offers a glimpse into the future of software itself. Open source doesn’t just benefit from community contributions that enhance functionality and scale distribution—it drives specialisation, customisation, and modularity. With AI further accelerating these trends, open-source frameworks are emerging as a more essential cornerstone than ever in the future of software development. If we’re heading toward an explosion of software applications, frameworks that abstract complexity and focus on core building blocks should become foundational. Much like the protocols that underpinned the internet, these open-source frameworks allow people to create bespoke applications from niche software to mainstream products.
From a cultural perspective, I would argue that software is evolving to resemble internet memes and other digital content we consume daily—constantly reused, remixed, and reimagined. Again, this is the benefit of open source, fostering broader developer participation, abstracting complexities, and continuously refining use cases. This creates a powerful flywheel: as more developers and builders contribute, the platform becomes more robust and versatile, attracting even more developers and expanding its potential. This is the commerce flywheel Medusa is building.
I highly recommend visiting their website, even if you have nothing to build. It’s a great piece of design. At its best, software is not just a technology; it’s a craft, and Medusa makes it beautiful.
I think we are overlooking the huge impact on consumer behavior and the adoption of generative AI for buying purposes.
Generative AI chats double traffic to retailers every two months, and consumers will start to expect AI capabilities to help them consume information and make buying decisions.
source: https://blog.adobe.com/en/publish/2025/03/17/adobe-analytics-traffic-to-us-retail-websites-from-generative-ai-sources-jumps-1200-percent)
I believe new AI-based search tools (Google AI Search, ChatGPT, Perplexity Shopping, etc.) are redefining the interface for information. They’re optimized for answers, making them ideal for research.
However, most consumer decisions are based on trust, identity, and emotional connection, not on raw information. Here lies the conversion gap between research and purchase.
For decades, brands have used static websites as their primary engagement layer. However, traditional stores aren't built to incorporate the real potential of generative AI capabilities.
That's why the future of commerce won't be either static stores or brands simply turning their websites into chatbots.
To stay relevant as a direct channel, brands must integrate generative capabilities into their core experience (their websites) to craft unique interfaces and content in real-time.
And yes, this is what we are doing at Enso. :)