Why more Amazon ad spend keeps eating your margin

Open an Amazon account that has been running paid ads for a year and you tend to see the same picture. Spend is up. Revenue has held or grown a little. And somewhere in the gap between the two, profit margins have shrunk. We see it often enough in the accounts we audit that it is worth saying plainly. On Amazon and Takealot right now, spending more is the easy part. Knowing whether that spend made you any money is the part many brands can’t answer.

Advertising on these platforms has become more expensive over the past year. ¹. Add the rising cost of fulfilment (FBA fees increased 12% in 2024 per Amazon’s guidance)² and the competition for Prime-eligible inventory, and the pressure on contribution margin is real. Most brand teams have felt it, but fewer have changed how they budget in response.

Visibility has become a rental expense

A few years ago a minority of Amazon sellers ran paid ads (around 40%). Today over 70% of sellers run ads ³. As more sellers bid for the same placements, organic results have been pushed down the page. If you are not advertising, a large share of buyers will not see you, so you advertise to stay where you already were.

Paying to hold your position buys you nothing new; it keeps you where you were, and the bill goes up every quarter.

Why ACoS on its own tells you almost nothing

Most brands set a target ACoS, watch the dashboard, and push spend up when sales dip. The trouble is that ACoS does not know your costs. A product running at 25% ACoS can be healthy or underwater depending on cost of goods, fulfilment, returns and your retail margin. Two products on the same ACoS can sit on opposite sides of profitable, and the dashboard won’t necessarily tell you which is which.

The brands that come through this period in good shape tend to share one habit. They know, per SKU, the most they can pay for a click and still make money, and they hold that line when the auction tempts them past it.

What profit-led advertising actually looks like

It starts before the budget.

Work out net margin per SKU after cost of goods, fulfilment fees, returns and any surcharges, then set your maximum viable cost per click from that. If you can’t afford today’s auction price on a product and still make money, the answer is rarely to bid more cleverly. Rather fix the economics of that product, or take it out of paid promotion.

Treat different traffic types differently.

Bidding on your brand name, generic category terms, and competitors’ names are three different jobs with different economics.

1. Defending your brand should be cheap and efficient.
2. Category terms are where growth money belongs.
3. Conquest is rarely worth it for a brand of modest size.

Blend all three into one campaign, the way most accounts are set up, and you lose the ability to see which of them is working.

And it means fixing the listing before you feed it traffic. A page that converts badly will convert badly no matter how many clicks you buy. Before lifting a budget, read your own listing as a buyer would.

• Is the main keyword in the title?
• Is the bullet copy about what the product does for someone, or only its specifications?
• Do the images show scale and context?
• Is there A+ content?

These move both organic rank and paid conversion, and improving them lowers your real cost per click without touching a single bid.

None of this shows up in a weekly ACoS report.

The question worth answering is not what you spent and what you sold. It is what that spend contributed to profitable revenue, and whether the same money would have done more elsewhere.

That means joining your ad data to your P&L rather than watching platform metrics on their own. It also means setting a floor as well as a ceiling: a return below which you pause and investigate, not only a monthly cap you spend towards. Retail media without a floor is a standing order to subsidise the platform’s growth with your own margin.

The South African angle

Amazon.co.za launched in May 2024, now about one year into local trading ⁴.

The novelty of being early has begun to wear off, and the question underneath it is tougher:

Is this channel actually profitable for us, or are we spending to look active? 

The same discipline applies to Takealot, where sponsored placements are expanding (brands see 258% average month-on-month growth after implementing Sponsored Advertising) and competition is rising, with Takealot revenue up 19% in 2025 to $706 million ⁵. As both marketplaces mature, the brands that build a profit-first framework now will be the ones that build a lasting advantage.

A marketplace is a commercial channel first, not a marketing one.

Run it like one: clear accountability, clean reporting, and every rand of spend treated as an investment you expect back.

If your Amazon or Takealot spend is climbing and you can’t say what it’s doing to your margin, an audit closes that gap.

At Ecommerce Counsel, we look at unit economics, campaign structure, and reporting together, and that’s where we’d start with you.