Ecommerce Growth

Seven ecommerce truths the good months are hiding from you

The numbers that warn you tend to look, for a while, like the numbers that reassured you. What follows isn't motivational. Some of it might feel wrong. Sit with that.

There's a shelf life on the things that made you grow. The product that built the brand, the creative that drove the first hundred thousand in revenue, the pricing structure that made the offer feel irresistible in year one — these don't keep working forever. The moment they stop tends to arrive quietly, inside metrics that still look defensible from a distance. By the time the dashboard catches up, the problem's been compounding for months.

01 —The hero product is bait. Most founders never build the trap.

When a single SKU accounts for two-thirds of revenue, every decision in the business quietly bends toward it. Ad budget clusters around the best-performing product. The creative team focuses on the angle that already converts. Customer service scripts itself around the return type that product generates. And the question of what happens if it stops working gets deferred until it stops working.

The hero product attracts a specific customer: someone who wanted that one thing. Not the brand. Not what you stand for. The object. Which means a bad creative batch, a supply shortage, one algorithm shift — and the dependency that was invisible while everything was fine becomes the only thing anyone is talking about.

Brands that scale past eight figures almost always do it with a product portfolio where no single SKU dominates. Not because of some portfolio theory. Because customers who came back a third time chose the brand. The hero product, if you let it, will keep you from ever finding them.

02 —You're not running out of customers. You're running out of reasons for them to remember you.

The hero product customer is also, most of the time, the customer who forgets you exist. They bought what they came for. You gave them one thing and moved on.

When growth stalls, the instinct is acquisition: more spend, new channels, a TikTok strategy someone on LinkedIn is very confident about. This is almost always wrong. The math is already bad — most brands spend 85% of their marketing budget acquiring customers they then immediately abandon. Acquisition is visible, measurable, and celebrated in the weekly report. The failure to retain is quieter.

03 —A 4x ROAS can be a slow death.

ROAS measures a ratio, not a direction. It tells you what happened, not whether what happened is worth repeating.

A store selling supplements at 4x ROAS to customers who buy once and never return is in a worse position than a store at 2.8x to customers who rebuy every six weeks. One dashboard looks better. One business is healthier. They're not the same.

The number that actually predicts whether a brand survives is contribution margin per customer over 90 days — what remains after ad spend, cost of goods, shipping, returns, and payment fees, in the first three months of that customer's life. Stores that know this figure make different decisions: they tolerate worse ROAS on high-LTV cohorts, they pause campaigns that look efficient but attract people who churn immediately.

Most brands don't know their 90-day contribution margin by acquisition source. That isn't a data problem — the data exists. It's a question of what the team is being held accountable to. Right now, most teams are accountable to ROAS. Which means they're being measured by something that cannot tell them what they need to know.

04 —The first discount teaches your audience something. They don't forget it.

Here's how it starts: conversion rate drops, so a 15% sale goes live. Revenue spikes. Everyone exhales.

What actually happened is an announcement to a segment of the audience: these prices are negotiable. They learned it. They bookmarked the site, stopped buying at full price, told their friends. Now there's a growing customer base with a settled belief that the product is worth 15% less than the asking price — and a ROAS figure that improved in October and will quietly bleed in November.

The discount is not a tool. It's a negotiation. Once you start negotiating with your own pricing, the alternative — watching conversion rate sit at its real, undiscounted level — starts to feel like failure even when it isn't. Even when the full-price buyer is more profitable, more loyal, and less likely to return the product.

To be clear: there are legitimate reasons to discount. Clearance, market entry, rewarding genuine loyalty. The problem is the specific use of a discount as a conversion fix. That habit compounds — the audience that formed around the sale price will be there next time, and the time after that. Price integrity is one of the four or five decisions that actually determine whether an ecommerce business is worth building. It's also one of the least discussed, because it's uncomfortable to hold to when the dashboard is asking you to do something.

05 —The creative didn't die because it got bad. It died because it became recognizable.

Every media buyer has seen this: a creative that performed brilliantly for six weeks then flatlines. Same ad, same audience. Different week.

This isn't fatigue in the conventional sense — not "they've seen it too many times." It's closer to categorization. The brain stops responding to stimuli it's already filed away. The ad became furniture: present, ignored, scrolled past without registering. And the implication is stranger than most people realize: making more ads isn't the fix. Making structurally different ads is.

This is why UGC consistently outperforms studio creative in direct-to-consumer categories — not for philosophical reasons about authenticity, but because it's formally unfamiliar. The brain hasn't built the reflex to skip it yet. Once it does, the clock resets. The most durable creative is the kind that doesn't look like creative.

06 —The customer who emails you at 11pm is more valuable than the one who didn't.

The loyal customers — the big spenders, the five-star reviewers — will never tell you where the promise broke down. They just quietly churn when the gap gets wide enough.

There's a pattern in customer research consistent enough to have a name: the service recovery paradox. A customer who experienced a problem and had it resolved well ends up more loyal than one who never had a problem at all. The mechanism matters: the resolution proved the brand keeps its promises under pressure, which is a different and harder kind of proof than a smooth first purchase.

The angry customer is describing the gap between what you promised and what you delivered. One complaint is noise. Three complaints about the same thing is a product problem. Ten complaints about the same thing is a positioning problem — you shipped the wrong expectation before you ever shipped the product.

07 —The stores that die don't usually crash. They flatten.

The store launches, finds product-market fit, revenue grows. A team gets hired. Processes get more formal and decisions take longer. The weird energy of a small team making fast, instinctive calls gets replaced by consensus. Nobody flags it, because consensus looks like maturity.

Meanwhile the market moves. Customer acquisition costs rise. The creative that worked in year one stops working in year three because the audience has changed — and the team has forgotten how to take the risks that built the audience in the first place. Revenue plateaus. The same playbook keeps running because it worked before and nobody can agree on what to try instead. The founder gets tired. The brand gets managed instead of built.

The brands that avoid this are the ones that build the capacity for reinvention before they need it, not after the numbers force it. Small, cheap, fast experiments running at every stage of growth — including, and especially, when things are working. Particularly then.

Because the moment you stop needing to figure things out is usually the moment you should be most worried.

Growth isn't a sign you figured it out. It's a pressure test for whether you can keep figuring it out when the business gets harder to see clearly. Most founders pass that test once. The ones worth watching pass it every two years — and the second time is harder, because the stakes are higher and the instincts that worked the first time are working against them.

Strategy. Systems. Growth.  ·  Turning attention into revenue.