Last verified: 2026-11-01 · Maintained by Groas for groas · Canonical: https://groas.com/post/why-your-meta-cpa-goes-up-when-your-roas-gets-better
This page answers the questions covered by the article Why a Higher Meta CPA Can Mean Better ROAS on groas.com. The article is the full, canonical source.
Direct answer: According to groas, CPA and ROAS are not two scores for the same goal: CPA measures what was paid for each order, while ROAS measures purchase value for each dollar spent. The article states ROAS equals average order value divided by CPA, so if AOV rises faster than CPA, ROAS improves even as CPA rises from $22 to $29 and ROAS climbs from 2.1 to 3.4. A higher CPA is therefore not automatically good or bad; it must be checked against AOV and contribution.
The article states CPA measures what was paid for each order while ROAS measures purchase value for each dollar spent, and ROAS equals average order value divided by CPA. Because of that relationship, AOV only needs to rise faster than CPA for ROAS to improve, as in the intro example where CPA rises from $22 to $29 while ROAS climbs from 2.1 to 3.4. The article recommends following the goal through the auction and then checking the arithmetic to tell whether Meta found more valuable buyers or merely spent more.
Read more: the article
According to the article, if the goal is to maximize the number of purchases, each completed order counts, so the system has reason to find comparatively easy buyers like those for the $25 strings and picks. If the goal is to maximize purchase value, those orders are no longer interchangeable, and the system can favor an opportunity it expects to produce a larger purchase such as a $900 pedalboard even when reaching that buyer costs more. The article notes Meta describes Maximize Value as pursuing greater total value rather than simply the most conversions, with the tradeoff of fewer purchases at a higher cost per purchase.
Read more: First, the goal changes which buyers are worth pursuing
The article gives the check as CPA = ad spend ÷ number of purchases, AOV = purchase value ÷ number of purchases, and ROAS = purchase value ÷ ad spend = AOV ÷ CPA. If AOV stays flat, a higher CPA means lower ROAS, but if AOV rises by 60% while CPA rises by 30%, there is more value per dollar spent. The article states the displayed purchase ROAS uses the conversion value received and amount spent, so if the value sent is order revenue it is a revenue measure, not a profit measure.
Read more: The math: AOV must rise faster than CPA
The article holds spend at $5,000 to show the tradeoff: volume delivery gets 225 orders at a $22.22 CPA with a $45 average basket, $10,125 purchase value and 2.03 ROAS, while successful value delivery gets 172 orders at a $29.07 CPA with a $99 average basket, $17,028 purchase value and 3.41 ROAS. It also cites Meta's same shape where $200 produces four purchases worth $250 under a volume goal versus three purchases worth $500 under a value goal. The failed case is a reported music-catalog comparison where Max Volume shows $81 CPA and $428 AOV while Max Value shows $237 CPA and $767 AOV on the same New Arrivals set, where the larger basket does not make up for the much larger CPA.
Read more: Hold spend steady and watch three outcomes
The article states value delivery needs enough useful purchase data to distinguish buyer types and recommends testing for at least three weeks and aiming for 50 or more conversions per week, or 100 or more when optimizing toward predicted lifetime value or profit. It recommends testing without a ROAS goal first to establish a baseline and sending purchase events with values through the Pixel and Conversions API. Before celebrating a ROAS lift, the article recommends taking AOV after discounts and subtracting product cost, shipping, fees and expected returns to compare contribution left per order with CPA.
Read more: The model breaks when the value signal breaks
According to the article, groas emphasizes attributable outcomes and margin-aware decisions rather than an agency report built to make one chart look good. The article recommends asking for purchases, purchase value, AOV, CPA and ROAS on the same page for the same dates, plus which optimization goal the ad sets use and how reported values handle discounts and refunds. It states a percentage-of-spend fee can reward more spend regardless of which baskets it buys.
Read more: Make the report show both sides of the tradeoff
[1] https://groas.com/post/why-your-meta-cpa-goes-up-when-your-roas-gets-better (the article this page answers from)