There's a moment in the maturation of any software category when the thing stops being a product and starts being plumbing. You know it's happened when the conversations shift from "should we use this?" to "how do we extend this?" and eventually to "wait, does anyone actually remember a time before this?" Multi-vendor marketplace infrastructure is in the early stages of that transition right now, and most people in commerce haven't clocked it yet.
When I was building recommendation and search systems at an Australian marketplace, the underlying platform infrastructure was a constraint you managed around. You wrote adapters for payment splits. You built custom tooling to handle multi-seller inventory states. You maintained parallel systems for seller onboarding because the core platform wasn't designed for the kind of dynamic catalogue that eight million monthly active buyers actually produce. That engineering overhead wasn't considered a strategic problem — it was just table stakes for running a marketplace at scale. Every team was building the same undifferentiated infrastructure. That's the signal that a category is about to become a platform.
What's shifted in the last few years is the emergence of marketplace SaaS platforms that treat multi-vendor complexity as the core design problem rather than an edge case. Companies like Marketplacer — which we backed in 2023 — are building infrastructure where seller onboarding, catalogue management, order routing, and commission accounting are first-class features, not bolted-on modules. The implication for anyone building a new marketplace is significant: the question is no longer whether to build or buy the marketplace layer, but how well the platform you choose handles the specific variant of multi-vendor complexity your market requires. That's a product question, not an engineering question, and it's pulling a lot of marketplace capability out of in-house engineering teams and into vertical SaaS.
We should be careful not to overstate where this transition is. The shift from product to infrastructure takes longer than the people selling the infrastructure would like to believe. There are real categories of marketplace complexity that current platforms don't handle well — highly regulated goods, cross-border tax treatment, complex service-layer marketplaces where the transaction isn't a widget but a time-based agreement. The companies that will define this category over the next decade are the ones that can extend their platform primitives into those harder problem spaces without losing the clean abstraction that makes the platform valuable in the first place. That's a hard technical and product challenge, and most platforms haven't had to face it at scale yet.
But the direction of travel is clear. In five years, the question won't be "how do we build a marketplace?" It will be "which infrastructure layer do we extend?" That's a materially different competitive dynamic for the companies trying to launch new marketplaces, and a genuinely interesting investment opportunity for those of us watching the infrastructure category take shape. We're early in the deployment phase of marketplace as infrastructure, and the companies that nail the platform abstraction problem in APAC's specific multi-currency, multi-corridor, multi-regulatory context will be in an extremely defensible position.