ACOS is ad spend divided by the sales those ads are credited with. It is the core efficiency number on Amazon and in retail media, and 8,100 people search the term every month in the US, per Semrush, with another 1,300 asking "what is acos" directly. Most of them get a formula. Few get the part that matters: what ACOS your brand can afford.
This is the fourth piece in our KPI cleanup series, after Modern KPIs: the rise of ACOS, MER and weighted CAC. Here we work ACOS from the unit economics up, using one $48 serum, then show how it changes for color cosmetics, replenishment skincare and medspa.
ACOS in one line
ACOS = ad spend ÷ ad attributed sales
Spend $4,800 on sponsored ads in a month. The platform credits those ads with $16,000 in sales. Your ACOS is $4,800 ÷ $16,000, or 30%. Put plainly, you paid $0.30 in advertising for each $1.00 of sales the ads were credited with.
Two words in that line carry the risk. Attributed means the platform decides which sales count, inside a click window it sets. Sales means revenue, not profit. ACOS never sees your cost of goods, your fees or your margin, which is why a 30% ACOS can be excellent for one product and a loss on another.
30%
ACOS on $4,800 spend
3.33
The same result as ROAS
12%
TACOS on $40,000 total
ACOS vs ROAS: the same ratio flipped
ACOS and ROAS measure one thing from opposite ends. ROAS is sales over spend. ACOS is spend over sales. A 30% ACOS is a 3.33 ROAS, a 25% ACOS is a 4.0 ROAS, and a 50% ACOS is a 2.0 ROAS. If your Meta team reports ROAS and your Amazon team reports ACOS, convert one before anyone compares them. Our ROAS calculator and our guide to return on ad spend cover the ROAS side.
The flip matters for one reason: direction. With ROAS, higher is better. With ACOS, lower is better. An ACOS moving from 22% to 31% plots as a line going up, which reads like growth. It is a 29% drop in efficiency.
Break even ACOS starts with unit economics
Your break even ACOS equals your margin before ad spend, expressed as a share of price. Above it, every ad sale loses money. Below it, every ad sale earns something. Work it out once per hero product.
WORKED EXAMPLE
One $48 serum, break even to target
- Price: $48.00
- Cost of goods at 20%: $9.60
- Marketplace referral fee at 15%: $7.20
- Fulfillment fee: $3.36
- Margin before ads: $27.84, which is 58% of price
Break even ACOS is 58%. Spend $0.58 to sell $1.00 of this serum and you make nothing. Set a 20% profit goal on each ad sale and your target ACOS is 58% minus 20%, or 38%.
Most brands never write this down. They inherit a target from an agency benchmark, often 25% or 30%, that has no link to their own margin. On the serum above, a 30% target gives up 8 points of headroom that could have bought more reach at 38%.
Target ACOS: choose the profit line on purpose
Target ACOS is break even ACOS minus the profit you want each ad sale to keep. That makes it a business decision, not a platform setting. Three common choices:
- Launch: target at or near break even, 58% on the serum, to buy reviews and rank while the listing is new.
- Scale: target 10 points under break even, 48%, to grow volume at a thin profit.
- Harvest: target 20 points under, 38%, once rank is established and the goal is cash.
Write the phase next to the target in the plan. An ACOS of 52% is a success in launch and a problem in harvest, and the number alone will not tell anyone which.
TACOS tells you if ads are building the brand
TACOS = ad spend ÷ total sales, paid and organic
ACOS only sees the sales the ads get credit for. TACOS, total advertising cost of sale, divides the same spend by every sale. In the month above, $4,800 of spend against $40,000 of total sales is a 12% TACOS.
The trend is the insight. Say three months later spend is $6,000, attributed sales are $18,750 and total sales are $60,000. ACOS rose from 30% to 32%. TACOS fell from 12% to 10%. Organic sales grew from $24,000 to $41,250, a 71.9% lift. Judged on ACOS alone, that quarter looks slightly worse. Judged on TACOS, it is the best quarter of the year.
Where ACOS misleads
- Attribution windows. The platform credits sales inside its own click window. A shopper who clicks, leaves and buys later outside that window never shows up, so ACOS reads high on products with long consideration.
- Brand search. Ads on your own brand name post low ACOS because those shoppers were already coming. Report branded and non branded ACOS separately or the average hides the cost of new demand.
- Halo. Amazon ads lift DTC and retail sales that ACOS cannot see, and Meta ads lift Amazon sales the same way. This is the gap TACOS and MER exist to close.
If your platforms report three different acquisition costs for the same month, the problem is attribution, not performance. Our 2026 skincare advertising benchmarks show how far platform reported CPA and ROAS drift across Meta, TikTok and Google.
WIN BEAUTY: color cosmetics break even lower than you think
Low price points compress break even ACOS because fulfillment fees are closer to fixed than proportional. Run a $22 lipstick through the same model: cost of goods at 20% is $4.40, the 15% referral fee is $3.30 and fulfillment is $3.22. Margin before ads is $11.08, or 50.4%.
That is almost 8 points less room than the $48 serum, on the same cost structure. With a 20% profit goal the lipstick target is 30.4%, against 38% for the serum. Bundle two shades into one $38 set and the fixed fee is spread across more revenue, which is the fastest lever on break even ACOS for makeup. Pair this with the Meta side in our breakdown of Meta advertising costs for beauty brands.
GROW SKINCARE: replenishment changes what you can afford
Skincare reorders. If 35% of new serum customers buy again within 90 days without further ad spend, each new customer is worth $27.84 in first order margin plus $9.74 in expected repeat margin, $37.58 in total. Against the $48 first order that is a 78.3% lifetime break even ACOS.
GROW SKINCARE
Two ACOS targets, not one
Set a first order target of 38% for all traffic and a new to brand target of 58%, lifetime break even minus the same 20% profit line. Only use the higher number where the platform reports new to brand sales, and only with a reorder rate you have measured, not one you hope for.
This is where ACOS connects to weighted CAC. Blending first order and lifetime targets into one figure hides which one is paying. For the full skincare picture see our beauty and skincare growth services.
GROW MEDSPA: ACOS by another name
Medspas do not sell through a marketplace, but the ratio still applies as ad cost to booked revenue. At a $90 cost per lead, a 45% show rate and a 50% consultation to treatment rate, acquiring one patient costs $400. Against a $1,200 first treatment, that is a 33.3% ad cost of sale. Set the same break even logic against treatment margin, not ticket price. Our medspa benchmarks cover cost per lead by channel, and the medspa growth page covers how we run it.
Where ACOS sits in the KPI stack
ACOS is a channel number. It should never be the number that sets your total budget. Use each metric at its own level:
- ACOS: one marketplace channel, set against product margin.
- TACOS: the marketplace as a whole, including organic lift.
- MER: the business. $160,000 of total revenue across DTC and Amazon on $44,800 of total marketing spend is a 3.57 MER.
- Weighted CAC: acquisition cost blended across channels, weighted by the new customers each one actually brought in.
Read them top down. If MER holds and ACOS rises, the channel may simply be taking on more of the acquisition work. If ACOS holds and MER falls, something outside the channel broke. The same cost logic runs through our CPM calculator, which is where media efficiency starts.