Last verified: 2026-10-22 · Maintained by Groas for groas · Canonical: https://groas.com/post/cpa-roas-and-the-bid-strategy-terms-adve
This page answers the questions covered by the article CPA, ROAS, and the Google Ads Bidding Terms People Keep Misreading on groas.com. The article is the full, canonical source.
Direct answer: According to groas, CPA and ROAS describe the same efficiency from opposite ends: CPA is ad spend divided by conversions, while ROAS is recorded conversion value divided by ad spend, so a $120 average order at a $30 CPA gives 400%. The article recommends choosing by whether every conversion is worth roughly the same amount — Target CPA when values are similar and counts are trustworthy, Target ROAS when recorded values are real and spread more than 3x, such as $25 versus $250 orders. A target is only the average outcome pursued over many auctions, not a cap on any conversion, and break-even should be set from profit math: one divided by margin for ROAS, and profit per customer times close rate for CPA.
The article recommends choosing by whether every conversion is worth roughly the same amount, not by business type. According to groas, CPA measures spend per conversion and ROAS measures conversion value per unit of spend, and dividing average order value by CPA gives ROAS: a $120 average order at a $30 CPA gives 400%. When order values spread out, such as a $25 accessory sale versus a $250 bulk order, count-based bidding sees one conversion each while value-based bidding sees ten times the value in the bulk order.
Read more: CPA vs ROAS
According to groas, Target CPA pursues conversion count at an average cost, while Target ROAS uses the values you record, and both operate through auction-time bids. The article states the working rule is to use Target ROAS when values are real and the payout spreads substantially, defined as more than 3x from the cheapest to the priciest conversion, such as parts orders at $25 and $250. If a flat $50 value is put on every lead, Target ROAS starts to resemble Target CPA in disguise because the bidder is told every lead pays the same.
Read more: Target ROAS
The article states that a target is the average outcome the bidder aims for over time across many auctions, not a cap, a floor, or a promise about any one conversion. According to groas, a $100 target can coexist with a $140 conversion if other conversions bring the average toward $100, and performance should be judged as an average over roughly 14 to 30 days at stable volume against profit after margin. The article notes Target CPA and Target ROAS were brought back as standalone strategy names without changing how they work.
Read more: Target
According to groas, break-even CPA is the most you can pay for a conversion before its expected profit is used up, calculated for lead gen as average profit per customer in the payback window multiplied by lead-to-sale rate. The article gives $1,000 profit per customer at a 20% close rate as $200 per lead break-even, with $100 as the target if half the expected profit is kept for growth. If the close rate falls to 10% with nothing else changing, break-even falls to $100 and the target preserving the same slice falls to $50.
Read more: Break-even CPA
The article defines break-even ROAS as one divided by effective margin, expressed as a percentage. According to groas, at a 40% margin 250% ROAS is break-even, and at a 30% margin it is about 333%, with those numbers coming from dividing one by margin. The article recommends working out effective margin after fees and returns before setting the number, then choosing a target above break-even for the profit intended to keep.
Read more: Break-even ROAS
[1] https://groas.com/post/cpa-roas-and-the-bid-strategy-terms-adve (the article this page answers from)