Ecommerce is retail.
Not retail-adjacent. Not a technology business that sells products. Retail - the act of buying and selling goods, through a digital channel. That is what it is, and the P&L of most ecommerce businesses operating today makes that case more clearly than any argument I could put forward.
I recognise this sounds reductive. The industry has spent considerable effort arguing otherwise. And for a long time, it wasn't wrong to.
The Case That Made Sense
Ecommerce grew up in proximity to technology and relied on it intensively - for listings, payments, logistics, advertising. That proximity shaped our identity. But so did the early growth rates. When customer acquisition was cheap, iteration was fast, and margins were generous, there was little pressure to develop the commercial rigour that leaner environments demand. We developed our own language - ROAS, LTV, CAC, conversion rate optimisation - aligned with technology conferences rather than retail trade shows, and built our own talent pipeline trained on ecommerce-specific frameworks.
More importantly, we developed our own economics. And for a meaningful period, those economics genuinely were different.
We had structural cost advantages that physical retail didn't. Smaller premises, lower rents and business rates, fewer staff per unit of revenue. Faster iteration cycles - where a physical retailer commits to a range 12 to 18 months before it reaches the shop floor, we could launch, test, and pivot within weeks.
These were real structural advantages. They produced real margin outperformance. And they justified treating ecommerce as a genuinely distinct operating category.
What Changed
Every one of those advantages has eroded or has been replaced by a different cost.
Smaller premises and lower rents? Replaced by platform commissions. Shopify, Amazon and TikTok spent years building global infrastructure and are now extracting a return on that investment through commissions and fulfillment fees that are permanent fixtures of the cost base.
Advertising is now a hefty pay-to-play digital shelf-space fee that scales with revenue. It has become a structural cost of trading online, not a discretionary marketing lever. Organic visibility still exists, but it is increasingly reinforced by paid activation; without paid support, organic reach often weakens too. The burden is not just spend, but the continuous management of algorithms, formats, creative volume and measurement rules.
The linear headcount model has been replaced by tech stack costs and agency dependency - costs that can scale faster and with less visibility than headcount ever did.
And faster iteration was eroded by supply chain complexity and lead times. When you add that ecommerce brands now need to be present and coordinated across multiple channels - because that is where the market is - the planning and operational complexity starts to look a lot like traditional retail. It is not just an expectation. It is a structural requirement of where the market has moved.
The result is margin convergence working through two distinct mechanisms. Above contribution margin, structural cost absorption - platform commissions, advertising, fulfilment - erodes margin directly and permanently. Below contribution margin, the operational complexity of running across multiple channels adds headcount and attention costs that traditional retail has always carried. We are now carrying them too.
Online retail businesses no longer operate with the margins of 10 years ago. The arbitrage we were running has been competed away - structurally, not cyclically. Those costs are now built into the infrastructure of every platform. They are not going back.
The arbitrage we were running has been competed away - structurally, not cyclically.
What the P&L Says
The shape of the P&L of a well-run ecommerce business today and that of a brand distributed through physical retailers are now remarkably similar. Gross margin: similar range. Contribution margin after variable costs: similar. Inventory risk profile: similar. Capital intensity as a function of revenue: similar.
Ecommerce is not a technology business. It is a retail business that uses technology. The technology has been commodified. The retail economics underneath it have not changed.
The Missing Layer
This does not mean our ecommerce-specific metrics are wrong. ROAS, CAC, LTV - these are still useful instruments. The problem is that they are channel-level metrics, and for most of the industry's development they were operating within a margin environment generous enough that the commercial layer above them didn't need to be precise.
That margin environment is gone. The commercial layer above the channel metrics - the one that governs whether the overall P&L is structurally sound - is what most ecommerce businesses are still underbuilding.
Retail calls this discipline commercial thinking. It means knowing precisely what is adding to your margin and what is eroding it - which channels, which SKUs, which territories, and which marketing activations are having what precise impact on your P&L. Which decisions are really capital allocation decisions wearing the costume of marketing decisions.
Commercial thinking is common vocabulary in mature retail. It is not yet common vocabulary in ecommerce. That is the gap - not in channel execution, which is generally strong, but in the commercial architecture above it.
Ecommerce leaders are often optimising channels when the real question is whether the P&L can carry the commercial strategy.
Once you frame ecommerce as retail, the questions that matter change. The issue is not only how to optimise ROAS, but whether the P&L is structurally sound. Not only which platform to expand to, but whether the margin architecture can support that expansion.
The channel is digital. The economics always were retail.
