From €5M to €100M+: what changes when an e-commerce category scales

Scale exposes the absence of a system

When a category is still small, individual effort can hide structural weaknesses. A strong local team, a favourable campaign or one successful product can create growth. But those ingredients do not automatically travel to the next market.

At Emma, the Beds opportunity was clear. The challenge was making it repeatable. Moving from roughly €5M towards more than €100M in category revenue required a shared way to decide what to launch, how to price it, how to validate demand and when to scale investment. The category could no longer depend on a few people carrying the logic in their heads.

Start with the minimum useful portfolio

My earlier instinct was to perfect the assortment before launch. In practice, waiting for complete knowledge would have made us slower without making us much smarter. We needed enough range to test the core hypotheses, not enough range to answer every imaginable customer preference.

We built the early proposition around three broad tiers: an accessible bed, a mid-tier minimalist option and a premium boxspring proposition. Neutral colours and selected supplier-catalogue or private-label products compressed development time. This was not a compromise on customer experience; it was a deliberate way to separate the questions that mattered from the details that could wait.

• Will customers buy a bed from us?
• Which price and benefit architecture creates trade-up?
• Which proposition travels across markets?
• Can sourcing, delivery and assembly economics support growth?

Standardise the learning, not every local answer

International scale is often framed as a choice between global standardisation and local autonomy. The more useful answer is to standardise the learning system. Markets can adapt price points, media, financing or channels, but they should not reinvent the question, the economics or the success criteria every time.

Each launch therefore needed a common hypothesis, a defined minimum assortment, contribution expectations, owners and a review cadence. Results from one market had to change the next launch. A market launch stopped being an isolated project and became another observation in a shared category model.

Revenue is not the scaling gate

A product can sell and still be the wrong product to scale. Beds carry acquisition costs, fulfilment complexity, delivery failure risk, returns, claims and working-capital requirements. At category scale, a small error in one of those assumptions becomes a large P&L problem.

The operating question was therefore not simply whether a launch generated revenue. It was whether customer demand, product margin, acquisition, operations and after-sales economics worked together. Performance marketing had to be accountable to contribution and payback, not only to topline or platform ROAS.

The organisation is part of the product

Once the model showed repeatability, the next constraint became organisational. Scaling required clearer principles, roles, workflows and decision rights across product, markets, sourcing, marketing and operations. We built a product-success organisation to carry that system and keep the feedback loop alive.

That remains one of my most important lessons from Emma: scale is rarely the result of one exceptional launch. It comes from building an operating system that makes the next launch less dependent on heroics and more likely to succeed.