Not the customer's moment of buying. The company's.
I have been selling in the Amazon ecosystem since 2014.
It is a brutal school.
Amazon punishes without asking questions. If the product is bad, it shows. If customer messages are not answered in time, it shows. If a product runs out of stock, it shows. If there is too much stock, that shows too, in money: from day 181 Amazon starts adding an aged inventory surcharge on top of the normal storage fee, a penalty for goods it has grown disappointed in. And from October to December the UK storage rate for standard-size goods nearly doubles, from £0.76 to £1.51 per cubic foot per month. Amazon's own price list, not my opinion. Too many returns? That shows in the margin, in account health and eventually in your nervous system.
But this is exactly why Amazon has been an excellent school.
It forces you to look at a product business as it actually is. Not as a brand, not as a webshop, not as sales. As goods, margin, stock, advertising, ranking, returns, payment terms and cash.
Everything is connected to everything.
On Amazon it is not enough that you can sell. You have to know what that sale is allowed to cost.
This is maybe the most important thing I have learnt in 12 years. In a physical product business demand does not yet mean good business. You have to meet the demand with the right product, the right price, the right stock, and a cost structure that leaves money over for you as well.
I am used to thinking about my business one product at a time.
For many years I ran an inventory of about 5,000 products, replenished weekly into three warehouses. It was a more cash-neutral model. Lot of products, constant topping up, constant watching, constant decisions.
Since then we have moved to our own brands and our own products.
The game changed.
When you buy thousands of units of one own product, the price of a mistake is in a different league. You are not topping up a shelf any more. You are tying up capital and taking a view on demand before any money has come in. And you pay for that view up front. The standard supplier terms are 30% on order and 70% before the container ships, so the entire purchase price has left your account before the goods reach British water. I have written that cycle down elsewhere: from the first payment to the first money back, roughly 16 weeks.
I know what a wrong view costs because I have taken one. We bought garden ornaments once. The research said there was a gap in the market, so we built a product with every feature in it, made variations of it, and bought them all at the same time, before the market had told us anything. The niche turned out to be browse a lot, convert low. High clicks, high advertising costs, a selling price nobody would pay. The stock went eventually, at prices I would rather not type out. The bigger cost was the months of team time spent designing, repricing and pushing a product that should never have been ordered in that quantity. Nobody refunds that.
And that is why the future of the company really is decided at the moment of buying.
We sell goods.
Nothing else.
If the goods run out, there is no revenue. If there is too much, the margin gets eaten by discounts, storage fees or rising advertising costs. If the goods are wrong, they take space, money, focus and eventually the courage to buy the right goods.
This is not a small leak either. IHL Group estimates that stock being wrong, too much of it or none of it, eats 6.5% of everything retail sells worldwide. And in UK, wholesale and retail is the second biggest sector for insolvencies: 3,463 companies in the 12 months to June 2026, 15% of all of them. The statistics do not say why each one failed. I have my guess about where lot of that money was standing.
So after all this, do I think inventory is the enemy?
No.
Bad inventory is the enemy.
Good stock is capital that moves. Goods that customers want, that turn, that produce margin and give the company permission to grow. Bad stock is the same capital stopped: pallets that sit, force discounts, raise advertising costs and make the founder careful at exactly the moment they should be brave.
Attention decides which one you end up with. If the attention is on revenue alone, the company starts chasing revenue. If the attention is on advertising, the company optimises advertising. But if nobody is watching where the money actually gets tied up, a company can grow in the wrong direction for a long time before anyone stops to ask what is actually being built here.
That is why ecommerce for me is not primarily a marketing game. It is a capital allocation game.
Where does the next pound get tied?
Into more of this product? A new launch? More advertising? Clearing the old stock at a discount? Or nowhere, because none of the options is good enough? Keeping the money in the bank is also a decision, and some months it is the best one available.
A bad decision does not stay in Excel. It comes into the warehouse and into the cash, and from there into the team's daily work and the founder's sleep.
After 12 years I am fairly convinced of this:
Profit is made at the moment of buying.
Not the customer's moment of buying.
The company's.
When you decide what to buy, how much, when, which demand to believe and how much cash to tie up. In that moment a surprisingly large share of the future result gets locked in.
Marketing can help. Sales can help. Brand can help.
But if the buying is wrong, everything else is working uphill.
Amazon UK monthly storage fees and the aged inventory surcharge from day 181: Amazon's own UK pricing pages, retrieved 12 August 2026. Inventory distortion at 6.5% of global retail sales: IHL Group, September 2025. UK insolvencies: The Insolvency Service, company insolvency statistics for the 12 months to June 2026, wholesale and retail trade 3,463 cases, 15% of the total. Supplier payment terms and the 16-week cycle: sourced in the earlier article on where the cash goes in a product business.