Harpy Glossary

CPS (Cost per Sale)

Amazon & D2C glossary · Harpy Media

CPS (Cost per Sale) is ad spend divided by the number of orders it produced — what each sale cost you in advertising. It’s the inverse view of conversion efficiency: where ACOS asks “what share of revenue went to ads?”, CPS asks “what flat price did I pay per order?”

What is CPS?

That flat frame makes CPS unusually good for planning: it compares cleanly across price points, plugs into breakeven math beside contribution per order, and — via its CPC÷CVR identity — exposes exactly which lever (cheaper clicks or better conversion) will move it fastest.

The two formulas (and what each teaches)

Direct: CPS = total spend ÷ total orders — the accounting view. Predictive: CPS = CPC ÷ conversion rate — the operating view. The second is the strategic one: it proves a seller can cut cost-per-sale two ways — pay less per click (bids, precision, relevance) or convert more of the clicks they already pay for (listing, price, reviews). Since conversion work compounds and bid work resets weekly, the formula quietly ranks the two strategies’ durability. It also prices keywords before you scale them: $1.20 CPC at 8% CVR predicts a $15 CPS — know that before the budget, not after.

CPS = spend ÷ orders = CPC ÷ CVR · Breakeven CPS = contribution per order ÷ target ad shareThe left formula reports; the right one decides.

Where CPS is the right lens (and where it isn’t)

CPS shines in planning and comparison: breakeven CPS (contribution per order ÷ target ad share) becomes a crisp campaign ceiling; cross-product comparisons aren’t distorted by price the way ACOS is. Its blind spots: it ignores order VALUE (a $9 CPS on a $200 AOV bundle and on a $25 single are different worlds) and it counts orders, not units or customers — multipack and repeat-purchase economics need CPPU and CAC respectively. Use CPS for order-level planning, ACOS for revenue-share management, and always know which lens you’re holding.

In practice

A brand budgets a product launch around predicted CPS: top keywords model at $1.10 CPC and 9% CVR → $12.20 expected CPS against $30 contribution per order — a green light with a real ceiling. Two weeks in, actual CPS runs $19: conversion is 6%, not 9%. The formula names the lever — listing work, not bid cuts — and a rebuilt main image and A+ lift CVR to 10%, dragging CPS to $11 without touching bids.

⚠️ Watch out. A team sees CPS rising and reaches for the only lever they know: cutting bids. Clicks get cheaper, traffic halves, and CPS barely moves — because conversion was the actual problem. They saved the wrong dollars and starved the campaign that was one listing fix from working.

💡 Harpy tip. Before scaling any keyword or campaign, predict its CPS (CPC ÷ expected CVR) and compare to breakeven. Thirty seconds of arithmetic prevents most scaling regrets.

How Harpy Media helps

We forecast cost-per-sale before budgets and diagnose against the CPC÷CVR identity when reality disagrees — the formula tells you which lever, we pull it.

CPS FAQ

What is Cost per Sale?

Ad spend divided by orders generated — the flat advertising price of each sale.

How is CPS different from ACOS?

ACOS is spend as a percentage of revenue; CPS is spend per order — price-blind and better for planning and cross-product comparison.

How do I lower my CPS?

Two levers, per the CPC÷CVR identity: cheaper clicks (bids, precision) or higher conversion (listing, price, reviews) — conversion work compounds longer.

Want these numbers watched for you, every week?

Book Free Consultation