Skip to content
Picaro.AI Picaro.AI Prompt Library

Why ACoS alone is not enough

The most widely used metric on the Amazon Ads floor is ACoS (advertising cost of sales). It is calculated as ad spend divided by ad-attributed revenue, and lower is generally treated as “more efficient”. Optimizing purely for ACoS, however, often ends up shrinking both top-line revenue and profit. This page is not an argument against ACoS — it is a map of what to read alongside ACoS so that the metric connects to business decisions rather than distorting them.

What ACoS measures and where it fits

Section titled “What ACoS measures and where it fits”

ACoS is defined as:

  • ACoS = ad spend ÷ ad-attributed revenue

It captures how much direct revenue each Sponsored Products (SP) campaign — and other ad campaigns — generated, which makes it well suited to efficiency comparisons at the campaign or keyword level. If two campaigns run on the same product with ACoS of 12% and 28%, the latter is the one with room for bid and match-type adjustments. That is the comparison ACoS was built for.

The structural issue is that ACoS only has ad-attributed revenue in the denominator. Advertising spills over into organic traffic and other channels, but none of that shows up in the formula. Four typical traps grow out of that blind spot.

Four traps of optimizing for ACoS alone

Section titled “Four traps of optimizing for ACoS alone”

1. Organic cannibalization

Section titled “1. Organic cannibalization”

When too much budget flows into brand or branded-intent keywords, revenue that would have come through organic search results gets booked as ad-attributed revenue instead. ACoS looks healthy on the surface, but the underlying reality is “ad spend that did not need to exist” piling up. When a brand campaign consistently runs under 5% ACoS, there is reason to suspect cannibalization.

2. Shrinking the business

Section titled “2. Shrinking the business”

The fastest way to push ACoS from 20% to 15% is to drop bids and reduce exposure. Lower bids mean competitors take the slot, ad-attributed revenue itself contracts, the pace of review acquisition that supports organic rank slows down, and six months later total revenue is lower. ACoS improves while the business shrinks.

3. Over-investment in defensive keywords

Section titled “3. Over-investment in defensive keywords”

ACoS is dramatically lower on branded-intent keywords. When optimization runs purely on ACoS, bids naturally concentrate on these defensive keywords, and exposure on offensive keywords (generic terms, competitor conquesting) wastes away. The acquisition channel for new customers quietly closes.

4. Blocking new product launches

Section titled “4. Blocking new product launches”

New products have few reviews and lower conversion rates, so initial ACoS in the 40–60% range is not unusual. Mechanical budget cuts based on an ACoS threshold strip exposure before the product can climb the rankings. High ACoS during launch is not a problem in itself, but it becomes one if ACoS is the only KPI.

What TACoS reveals

Section titled “What TACoS reveals”

The metric to read alongside ACoS is TACoS (Total ACoS).

  • TACoS = ad spend ÷ total revenue (ad-attributed + organic)

Where ACoS measures “advertising efficiency”, TACoS measures how much of total revenue depends on advertising. If ad spend rises and organic revenue rises with it, ACoS may stay flat while TACoS falls. Conversely, if ACoS falls but TACoS gets worse, trap 1 (organic cannibalization) or trap 2 (shrinking the business) is likely in play.

TACoS is also the easier metric for executive conversations. It frames advertising as “how much the ad investment is moving the entire business”, which is the same language as the promotional-cost ratio on the P&L. The reason Picaro shows ACoS and TACoS side by side in daily reports is that the picture changes dramatically when only one of them is in view.

A multi-dimensional read with ROAS, SQP, and search share

Section titled “A multi-dimensional read with ROAS, SQP, and search share”

ACoS and TACoS together already widen the field of view, but three more metrics are needed to make the read reliable.

  • ROAS (ad revenue ÷ ad spend): close to the inverse of ACoS, but framed in profit terms as “how many dollars of return each dollar of ad spend produces”. Easier to use when aligning internal stakeholders.
  • SQP (Search Query Performance): for each search query, what share of the category’s impressions, clicks, and purchases the brand captured. ACoS lives inside the campaign; SQP shows the brand’s position within the market.
  • Search share (impression share / search query share): how often the brand appears on key terms. A drop in defensive-keyword share is an early signal that a competitor has started conquesting.

The same “ACoS 12%” reads very differently depending on context. If SQP shows purchase share falling from 30% to 18%, that 12% may mean the brand is preserving ad efficiency while losing market position. If SQP share is climbing, the same ACoS reads as a healthy offensive result. The interpretation flips with the surrounding metrics.

The role of label design

Section titled “The role of label design”

Multi-dimensional reading requires keywords to be grouped into meaningful units in advance. The five-axis label design Picaro recommends:

  • Offensive keywords: generic terms, use-case terms, need-state terms. The entry point for new customers.
  • Defensive keywords: brand names, product names, branded-intent terms. ACoS is low here, but this is also where over-investment risk lives.
  • Competitor: competitor brand names and competitor ASINs used for conquesting.
  • brand: the brand’s own name itself (the most branded-intent band within defensive keywords).
  • use: use-case and scene-of-use terms (“for camping”, “for nighttime”, etc.).

With these labels in place, structures like “ACoS looks fine, but spend is skewed toward defensive keywords” or “search share on offensive keywords is falling” become visible. Without labels, only the overall average ACoS is visible — and trap 3 (over-investment in defensive keywords) can never be detected.

Two case studies

Section titled “Two case studies”

Case A: “ACoS improved, but the business shrank”

Section titled “Case A: “ACoS improved, but the business shrank””

A health-foods account, ¥2.8M (JPY) monthly ad spend. ACoS improved from 22% to 16%, but total revenue shrank from ¥18.5M (JPY) to ¥15.4M (JPY) over the same period. TACoS got worse, from 15.1% to 18.2%. SQP showed that purchase share on a key generic term (“supplement recommended”) fell from 11% to 6%. Bids on offensive keywords had been cut too aggressively, and competitors took the market. As a single number, ACoS “improved” — in the multi-dimensional read, this is a regression.

Case B: “ACoS got worse, but it was the right investment call”

Section titled “Case B: “ACoS got worse, but it was the right investment call””

A kitchenware new product, ASIN B0XXXXXXXX, launch phase. First-month ACoS was 48% (¥0.95M (JPY) ad spend, ¥1.98M (JPY) ad revenue). On ACoS alone the budget would be cut immediately, but TACoS was 22%, with organic revenue ramping up in step with ads. SQP showed click share on key terms rising from 0% to 8% — the product was beginning to register on search results pages. Three months later ACoS landed at 24% and TACoS at 9%. The initial high ACoS was a rational launch cost in retrospect.

Summary

Section titled “Summary”

ACoS is still a valid metric for measuring “ad campaign efficiency”, but on its own it does not speak the language of business decisions. Use TACoS to read ad dependency, ROAS to align return at the granularity executives expect, SQP and search share to read market position, and labels to separate out “which type of keyword is doing what”. Reading these metrics together, rather than narrowing down to one, is the shortest path around the ACoS traps.

What to do next

Section titled “What to do next”