โ† Back to Blogs

Amazon Ads

9 min read

Your First-Order ACoS Cap Is Paying a Competitor To Keep Your Customer

Onieque Edwards

Content Strategist /Blog Writer

A supplement brand sets a rule for its Amazon ads. No campaign runs above the ACoS where the first sale breaks even.

It sounds disciplined. Every ad dollar has to pay for itself on the order it produces. Nothing runs at a loss.

For a product people buy once, that rule is fine. For a product people run out of and buy again, it is quietly one of the most expensive decisions in the account.

The cap prices every new customer as if they will buy exactly once. If the typical customer comes back four more times in the year, the brand is refusing to pay more than one sale's worth of margin to win a customer worth several. And the shopper does not disappear when you lose the click. They buy from whoever won it, and the reorders go there too.

What a first-order cap is actually assuming

ACoS is ad spend divided by the sales those ads produced. It is a useful number. It just answers a narrow question: did the ad pay for itself on the order it was credited with?

For a durable product, that is close to the whole question. Someone buys a blender, and the brand may not see them again for years.

For a consumable, it is the first page of the story. A customer who finds your magnesium through an ad and likes it does not search for magnesium again next month. They tap Buy Again, or set up Subscribe & Save, or search your brand name. None of those later orders are credited to the ad that found them. But none of them happen without it.

So a first-order cap is making an assumption it never states out loud: that a customer is worth one order. For most consumable brands that is wrong, and it is wrong in the same direction every time.

What a customer is worth over a year

Here is a worked example. The numbers are round on purpose. They are not from an account, and yours will differ.

A supplement sells for $30. After the referral fee, the FBA fee and the cost of the goods, say $14 is left from each order before advertising.

A first-order cap at break-even says the most you can spend to win that customer is about $14 in ad cost. A tighter cap, which most brands run, says less.

Now say a typical customer buys five times in twelve months: the first order and four reorders. Some of those reorders go through Subscribe & Save at a discount, so take a little off each. Call the twelve-month margin from that customer around $64.

That customer is worth about five times the first sale, and the cap is pricing them at one.

The brand that knows this can pay well over $14 to win the first order and still make money on the customer. The brand that does not will be outbid on exactly the shoppers who were going to become its best customers, and it will read the lost impressions as a sign the term was too expensive.

Build the number on margin, not revenue

The fix is not to stop measuring. It is to measure the right unit: margin per customer over a period long enough to include their reorders.

For each hero product, the build is:

Orders per customer over 12 months, multiplied by what each order leaves after costs: the price, less the referral fee, the FBA fee, the cost of goods, any Subscribe & Save discount, and the ad cost you allocate per order.

Where does orders per customer come from? If the brand is registered, the Repeat Purchase Behavior report in Brand Analytics shows orders, unique customers and repeat customers by brand and by ASIN. Orders divided by unique customers, over a long window, gives a working estimate. If the ads account has Amazon Marketing Cloud, it can measure lifetime value and the time between purchases directly, and Amazon has extended the purchase history AMC can draw on from 13 months to five years.

Two rules keep this number honest.

Use margin, never revenue. A revenue-based lifetime value makes almost any acquisition cost look affordable. A customer worth $150 in sales sounds like it justifies a $60 acquisition cost. If they are worth $64 in margin, it does not. The revenue version is the opposite mistake to the first-order cap, and it is the one that burns cash faster.

Leave a buffer. Setting the target acquisition cost at the full 12-month margin means you only break even after a year, and only if the estimate was right. A working range is somewhere between half and two thirds of 12-month margin, closer to half if cash is tight or the data is thin.

In the example, that puts the target between about $32 and $43 to win a new customer. That is more than double the first-order cap, and still leaves margin on the table at the end of the year.

Which spend the target applies to

This is where most lifetime-value advice goes wrong, and where it can do real damage.

The longer target applies only to spend that wins new customers. It does not apply to spend that re-buys customers you already have.

A shopper who searches your brand name, or clicks a retargeting ad after buying from you last month, may have reordered anyway. Paying for that click is not acquiring a customer. It is paying again for one you already won. If that spend is judged on lifetime value, the account will happily overpay to re-sell to its own loyal buyers and report it as growth.

Amazon gives you the line to draw. It counts an order as new-to-brand when the shopper has not bought from the brand in the previous 12 months, and Sponsored Brands reports new-to-brand orders and sales directly. So the account splits in two:

  • Acquisition spend, on category and competitor searches where new-to-brand share is high, is judged on 12-month margin per new customer.

  • Retention spend, on branded search and retargeting past buyers, is judged on a much tighter return, because most of those buyers were already coming back.

An agency that reports ACoS and TACoS but never new-to-brand share cannot tell you which of these your money is doing. That is usually the first thing worth finding out.

What has to be true before you raise a bid

Raising targets on a lifetime value you have not checked is just a more sophisticated way to overspend. Four things need to hold first.

The repeat rate is measured, not assumed. Work out how long one unit lasts a customer, say 30 days for a month's supply, and judge repeat behaviour over a window at least twice that long. A product with a 60-day cycle looks like it has almost no repeat customers in a one-month view, even when loyalty is strong. Read brand-level and ASIN-level repeat rates separately, because a customer who buys a different product from the same brand is a repeat brand customer but not a repeat ASIN customer, and the brand number will always look better.

Stock holds. If inventory runs out, Subscribe & Save deliveries are skipped and subscribers cancel. The Subscribe & Save dashboard shows the share of scheduled units not delivered because of stockouts. A lifetime value that assumes four reorders is fiction if the product is out of stock for one of them.

Discounts do not stack below the margin. Subscribe & Save discounts, coupons and promotions can all land on the same order. If the deepest discount is going to a loyal customer who would have reordered at full price, it costs margin and adds nothing. Your margin per order has to reflect the discounts you actually give, not the list price.

Returns are counted. Amazon's repeat purchase revenue does not net out returned goods. In a category with meaningful returns, reorder revenue is overstated until you adjust for it.

If you cannot confirm one of these, the target waits. An unchecked lifetime value is not a conservative estimate. It is a guess with a bid attached.

The objection worth taking seriously

"Didn't those customers come back because the product is good, not because of the ad?"

Partly, yes. The ad did not cause the reorders. The product did. But the ad caused the customer to exist. A shopper who never tried your magnesium cannot reorder it, and the shopper who tried a competitor's first now has that brand in their Buy Again list instead of yours.

That is why the long target belongs only to new-to-brand spend. It is paying for access to a customer, not for credit on their reorders.

The harder question is whether a rise in repeat customers after an increase in spend was caused by the spend. To claim that, compare customers acquired in the higher-spend months with earlier groups, and rule out the other things that move repeat rates: a price change, a competitor stocking out, a new listing, a change in the Subscribe & Save discount. If you cannot separate them, report the change as associated with the ads, not caused by them.

Where the money should go

The decision here is about allocation.

A consumable brand on a first-order cap spreads its budget by what breaks even today. That pushes money towards branded searches and existing customers, which look efficient because many of them were buying anyway, and away from the category searches where new customers are actually won.

A brand that knows its 12-month margin per customer does the opposite. It knows which products bring in customers who come back, the gateway products, and it is willing to pay more on those to win the first order, because it knows what the customer is worth by month twelve. It holds retention spend to a tight return, so it is not paying twice for the same buyer.

The brands outbidding you on your category's best searches are not always overspending. Some of them know their numbers and you do not.

If your best customers buy five times a year, what is your cap actually telling you to pay for one?

FAQ

Is break-even ACoS too strict for consumable products?

Often, yes. Break-even ACoS judges an ad on the first order alone. For a product customers reorder, the first order is a small part of what the customer is worth, so a break-even cap underprices exactly the customers most likely to come back.

How do I work out customer lifetime value on Amazon?

Multiply orders per customer over 12 months by the margin left on each order after the referral fee, the FBA fee, the cost of goods, any Subscribe & Save discount and allocated ad cost. If you are brand registered, the Repeat Purchase Behavior report in Brand Analytics gives orders and unique customers. If you have Amazon Marketing Cloud, it can measure lifetime value directly.

How much should I pay to acquire a new customer on Amazon?

Set it from 12-month margin per customer, not revenue, and leave a buffer. A working range is half to two thirds of 12-month margin, nearer half if cash is tight or the repeat data is thin.

What does new-to-brand mean in Amazon Ads?

Amazon counts a shopper as new to brand if they have not bought from the brand in the previous 12 months. Sponsored Brands reports new-to-brand orders and sales, which shows how much of your ad spend is winning new customers rather than re-selling to existing ones.

Should branded search be judged on lifetime value?

No. Most shoppers searching your brand name already know you, and many would have reordered anyway. Judge branded search and retargeting on a tighter return, and keep lifetime-value targets for spend that wins new customers.

Does Subscribe & Save subscriber count tell me my repeat revenue?

No. Subscribers skip deliveries, change frequency and cancel, so subscribers multiplied by price is not recurring revenue. Many repeat orders also come from outside Subscribe & Save, through Buy Again, brand search and reorders. Read the Subscribe & Save dashboard alongside the Repeat Purchase Behavior report.

Sources

Amazon Ads defines new-to-brand shoppers as those who have not completed a conversion with a brand in the last 12 months, in its announcement of the long-term sales metric, which estimates the sales value a brand can expect over the next 12 months from ad-driven new customers. Amazon's Sponsored Brands guide covers the new-to-brand metrics it reports. Amazon's note on expanded AMC measurement describes the move from 13 months to five years of purchase history, and its Ad Intel piece on measuring awareness with AMC recommends measuring new-to-brand lifetime value and shifting from ACoS towards cost per acquisition. Amazon's overview of Brand Analytics lists the reports available to brand-registered sellers, including Repeat Purchase Behavior.

Onieque Edwards

Content Strategist /Blog Writer

Onieque is the brain behind bold Amazon growth strategies and structured business execution. He enjoys turning scattered ideas into clear, actionable systems that actually drive results. When heโ€™s not building out growth plans or refining campaigns, youโ€™ll likely find him exploring new coffee spots or getting lost in ideas that connect strategy with creativity.