How we'd fix a high ACoS beauty brand on Amazon in 90 days

The starting point

Picture a skincare brand, a made-up composite rather than any real company: a handful of serums, solid reviews and an ads bill that keeps growing faster than profit. For the worked example, its main product is a $36 peptide serum.

The unit math, all illustrative: $36.00 price, minus $9.00 product cost, minus $10.80 in Amazon fees (assumed at 30% of price, so swap in your real fees) leaves $16.20 before advertising. Break-even ACoS is $16.20 divided by $36.00, or 45%. You can check your own figure with our break-even ACoS calculator.

The account, also illustrative: 2,500 units a month is $90,000 in sales (2,500 x $36). Ad spend is $19,800 and ad-attributed sales are $36,000 from 1,000 units, so ACoS is 55% ($19,800 / $36,000) and TACoS is 22% ($19,800 / $90,000). Contribution after ads is $20,700 (2,500 x $16.20 = $40,500, minus $19,800).

Now split that ACoS into its two parts. At a $1.65 cost per click, $19,800 buys 12,000 clicks. Those clicks produce 1,000 orders, an 8.3% conversion rate (1,000 / 12,000). Each ad order costs $19.80 ($1.65 / 0.0833).

What is holding it back

1. $3.13Q2 beauty ROAS, the lowest category in Pacvue's report

   Start with the category, because demand is not the problem. Front Row's Q1 2026 snapshot estimates Amazon beauty at $8.1 billion, up 13%, with skin care at $2.7 billion and up 18%. Yet the auction is unforgiving. Pacvue's mid-year benchmarks, as reported by beBOLD Digital, put Beauty and Personal Care Sponsored Products at $3.13 ROAS in Q2 2026, the lowest category in that report. That works out to about a 32% ACoS (1 / 3.13), while Eightx lists 24% for beauty from AdBadger data and treats it as a plus or minus 3 point band. The two sets disagree, so we would ignore both as targets and set goals from margin. See what a good ACoS really means.

2. +36,280%copper peptide searches, per Front Row, pulling money into ingredient terms

   Second, hot ingredient terms pull in money. Front Row reports copper peptide searches up 36,280% and says Medicube grew 217% to an estimated $108 million in the quarter. We would expect well funded rivals on those terms, though that is our reading, not a sourced auction figure.

3. 8.3%conversion in this example, so each click brings only $3.00 in sales

   Third, the click is not the real leak. Eightx shows a $1.18 beauty cost per click and a 12% to 15% conversion range, though it notes that range is platform-wide and not measured for beauty alone. Even so, 8.3% looks weak. At 8.3% conversion, each click is worth $3.00 in sales ($36 x 0.0833).

4. −$3.60on each ad sale at 55% ACoS in this example

   Fourth, ads are buying sales at a loss. At 55% ACoS, each ad-attributed unit carries $19.80 of ad cost against $16.20 of contribution, a loss of $3.60. The most this brand can pay per click at break-even is price x conversion x break-even ACoS, or $1.35 ($36 x 0.0833 x 0.45). It is paying $1.65.

The 90-day plan

Nothing would be paused until its replacement has data, so sales are protected during the handover. Cutting bids first would only hide a page that does not convert.

Budget split

This plan would spend less, not more. Cutting the $3,960 leaves $15,840, and we would round the budget down to $15,750 so that every dollar has to earn its place. The illustrative split, which adds to $15,750:

Target metrics

These are day-90 goals, not promises, and every number is illustrative. At $1.40 a click, $15,750 buys 11,250 clicks. At 11.1% conversion that is 1,250 orders, a cost of $12.60 each, or 35% ACoS ($12.60 / $36). Ad sales would reach $45,000 (1,250 x $36). Organic units move from 1,500 to 1,550, for 2,800 total. Sales reach $100,800 (2,800 x $36), up 12%, and TACoS falls to 15.6% ($15,750 / $100,800).

At 35% ACoS an ad-attributed unit earns $3.60 ($16.20 minus $12.60), the exact reverse of today's $3.60 loss.

Now the honest part. Contribution would be $45,360 (2,800 x $16.20) minus $15,750 in ads, or $29,610. If 400 units are Subscribe and Save at a 5% seller-funded discount, that costs $720 (400 x $1.80), leaving $28,890. That is an $8,190 monthly lift over $20,700: $4,050 from spending less, $4,860 from 300 more units, minus the $720 discount. PPC Spider's full-service plan is a flat $1,500 a month for up to 25 ASINs and $20k in ad spend, so the net goal is a $6,690 lift. If new creative is needed, a PDP package from $1,050 per product on two products adds $2,100 once, which that lift would repay inside a month at goal pace. Plans are flat retainers with a 90-day initial term, and pricing is public.

What could go wrong

• Rank dip

  Lower spend could cost some organic rank on ingredient terms. Organic share would be watched weekly.

• Conversion stalls

  If the page stays at 10.4%, the $1.40 ceiling drops to $1.31 ($36 x 0.104 x 0.35) and fewer clicks would be affordable.

• Compliance

  Pattern says facility registrations renew every two years and that a lapsed one puts listings at direct risk. It also reports the FDA said on September 9, 2026 that it issues no registration certificates, so any vendor selling one deserves doubt. Check documents against the live listing before you scale.

• Dated stock

  Serums can fall under Amazon's expiration-dated rules, and a cut in ad spend slows sell-through. Confirm current requirements in Seller Central and review FBA requirements before a deep restock.

• Trend risk

  Copper peptide demand can fade as quickly as it rose. A brand should bid on an ingredient only if the formula truly contains it and the label says so.

When this plan is the wrong fit

This plan assumes a page that can be improved and a margin with room in it. If your conversion is already strong and you still run above break-even, the fix is price, pack size or cost. If your reviews are the real objection, ads and images will not solve it. In a launch, a high ACoS may be a deliberate cost of rank, so a cut could do harm. Makeup also differs: Circana reports prestige makeup up 3% with softer units in the first half of 2026, against prestige skin care up 9%, so demand itself may be the limit. For very small catalogs, a one-time listing rewrite may be the better first step.

Steps you can take this week

A brand owner or ads manager could finish these in an afternoon.

1. Finance

   Calculate contribution per unit and break-even ACoS for your top five products.

   Products running above break-even
2. Ads manager

   In the campaign manager, export 90 days of cost per click and conversion by campaign.

   Which of the two moved most
3. Ads manager

   In the Sponsored Products search term report, sum 60 days of spend with zero orders.

   Wasted spend as a share of total
4. Brand lead

   In Brand Analytics, compare top ingredient and benefit terms with your title and bullets.

   Top terms missing from the copy
5. Compliance lead

   Match each ASIN's facility, ingredient list and claims against packaging and FDA filings.

   Asins with a mismatch

Sources

• Retail Dive: Fragrance drives beauty sales as consumers seek 'emotional and functional value', Circana data, August 24, 2026 (retaildive.com/news/fragrance-beauty-sales-growth-first-half-2026/828479/)
• Front Row: The 2026 Amazon Beauty Snapshot, May 27, 2026 (frontrowgroup.com/insight/the-2026-amazon-beauty-snapshot-what-the-category-is-telling-you/)
• beBOLD Digital: Amazon's Q2 2026 Retail Media Benchmarks, reporting Pacvue data, August 12, 2026 (bebolddigital.com/news/amazon-q2-2026-retail-media-benchmarks)
• Eightx: Average Amazon ACoS by vertical 2026, May 29, 2026 (eightx.co/blog/average-ecommerce-acos-amazon-by-vertical-2026)
• Pattern: MoCRA for Beauty Brands Selling on Amazon, September 28, 2026 (pattern.com/blog/mocra-for-beauty-brands-selling-on-amazon)