Why Are Your Ecommerce Margins Shrinking? The Operational Diagnosis

Margin never collapses. It erodes. A point here from a freight surcharge, a point there from a 3PL renewal, half a point from the clearance sale that cleared last season's buying miss, and three years later a business doing twice the revenue is keeping less money than it did at half the size.

The reflex is to blame acquisition. Ad costs did rise, and we have written about what rising CAC actually meanselsewhere. But the acquisition story is half the diagnosis at best. The other half is operational, it compounds quietly, and almost nobody audits it, because no single line on the P&L ever looks bad enough to investigate. The emblem of the whole pattern is Gymshark: a pound kept in ten in 2020, a penny kept now, while revenue rose 148%. Growth did not protect the margin. Growth hid where it was going.

This is the diagnosis: the five operational leaks that shrink DTC margins, the tell for each, and the order to work through them.

First, stop looking at the blended number

You cannot fix a margin you can only see in total. The blended gross margin on the P&L is the average of every SKU, channel and mistake in the business, and averages are where problems hide. The diagnosis starts by rebuilding contribution per SKU, fully loaded: selling price, minus true landed cost, minus fulfilment cost per order, minus the payment, returns and markdown costs that attach to it. Do that for the top fifty SKUs and the shrinking margin stops being a mystery and becomes a list.

Then bridge it. Take the margin from twelve months ago and today's, and allocate the gap line by line: how much went to price and mix, how much to product cost, how much to freight, fulfilment, markdown, returns. The bridge is an afternoon of work and it converts an argument about whose fault the margin is into a ranked table of where it went.

The five leaks

Landed cost creep. The product cost on the supplier invoice held steady, so nobody looked further. But landed cost, the all-in figure with freight, duty and handling loaded, drifted. The usual driver is freight mode: the share of goods arriving by air creeps up year on year as planning slips, and every airborne kilo carries a multiple of the sea rate. Duty changes and currency drift stack on top. The tell: you cannot state the true landed cost of your top ten SKUs without opening a spreadsheet built last year. The fix starts with negotiating freight properly and understanding the real air versus sea premium, and the arithmetic on your own numbers takes minutes in the landed cost calculator in our tools hub.

Fulfilment cost creep. 3PL contracts do not get more expensive in one dramatic renewal. They get more expensive line by line: a storage escalator here, a peak surcharge there, a returns processing fee that quietly doubled. The tell: cost per order rising while average order value stays flat, and an invoice nobody reconciles line by line. The signs it is time to leave your 3PL are usually visible two renewals before anyone acts on them.

Markdown and dead stock. Every buying miss eventually reappears on the P&L wearing a discount. The clearance sale that moved last season's overbuy did not create the loss, it recognised it, and the storage paid in between made it worse. The tell: a rising share of units sold below full price, and a warehouse holding stock older than two seasons. This is the leak an inventory audit quantifies first, because the dead stock number funds the fix for everything else.

Stockout substitution. When the bestseller runs dry, the margin damage arrives twice: the sales lost while it was dark, and the premium paid to patch it, emergency air freight or dropshipping at a far higher unit cost. Both are consequences of the same planning gap, and we broke the full mechanics down in how to reduce stockouts. The tell: air freight and expedite fees that spike in the weeks after every launch.

Mix drift and discount dependence. The subtlest leak. The catalogue's centre of gravity shifts toward lower-margin heroes, bundles built for conversion rather than contribution, and sitewide promotions that started as tactics and became structure. The tell: gross margin percentage looks stable while contribution per order falls, or revenue grows while cash stubbornly does not.

The order matters

Work the list in the sequence above, not in the order of annoyance. Landed cost and fulfilment come first because they reprice every unit you will ever ship. Markdown and stockouts come next because they fund and stabilise the buying. Mix comes last because you cannot judge the catalogue until the costs underneath it are true. Along the way, benchmark yourself honestly: our brand benchmarks rank the DTC brands we cover by what they actually keep per pound of sales, from 33p at the top to negative at the bottom, and the spread is almost entirely operational discipline, not category luck.

When it is not operations

Honesty requires the other half. Sometimes the margin problem really is pricing power in a commoditising category, or acquisition costs structurally rising faster than lifetime value, and no amount of freight negotiation fixes a product people will only buy at a discount. The bridge exercise settles this without an argument: if the twelve-month gap allocates mostly to price and CAC rather than to cost lines, the fix is brand and offer, not operations. Most of the time, in our audits, it allocates to both, and the operational half is the faster half to recover, because it does not require a single customer to change their mind.

Frequently asked questions

What is a good net margin for a DTC brand?

There is no universal figure, and the public data proves it: the brands in our benchmarks range from keeping 33p per pound of sales to keeping nothing at all. The useful question is not the industry average but your own trajectory, what you kept per pound three years ago versus today, and where the difference went.

How do I know if it is CAC or operations?

Run the margin bridge. Allocate the twelve-month margin gap line by line and the split declares itself. As a rough tell before the analysis: if margin fell while conversion and repeat rates held, look at costs first.

How fast can operational margin recover?

Faster than most founders expect, because most of the leaks are contract and planning problems rather than market problems. Freight mode shifts and 3PL repricing typically land within one to two quarters. Markdown recovery takes a season, because the stock already bought has to clear first.

Should we cut costs or raise prices?

Diagnose before either. Price rises on top of a leaking cost base hand the gain straight back, and cost cuts chosen without the per-SKU view usually cut muscle. The bridge tells you which levers are actually yours.

Relevant Case Study

Dog Friendly Co

Dog Friendly Co grew fast enough that the supply chain underneath started costing serious money, freight, fulfilment and inventory all leaking at once. One audit found seven figures of it, and the engagement that followed turned the findings into banked savings.

$1m +

Identified in an audit

$390k

Savings in the first 4 months