Meta Ads ROAS dropped: what changed and what to check
Check how Meta counts, whether the ad is tiring, whether prices rose, and whether the calendar or tracking moved. Meta's own results say its average price per ad rose year over year in the second quarter.
Three platforms, three answers, one budget to defend. Your GA4 export already holds the answer, and the first finding is free and never leaves your browser. Get your first finding free
By Joris van Huët, Founder & CEOUpdated 5 min read
Run the numbers for your store: the free break-even ROAS calculator, or the free safety stock calculator.
Before you change a creative, check whether Meta changed what it counts and what it charges. Meta's second-quarter 2026 release says its average price per ad rose 12% year over year while ad impressions rose 14%. For illustration: if your price per thousand impressions rose 12% and nothing else moved, a flat budget buys 1 ÷ 1.12 = 89.3% as many impressions, and a ROAS of 3.00 becomes 2.68. Work through five checks in order: counting, creative, price, calendar and tracking, then whether the drop is real.
Did Meta change how it counts your results?
Meta's Insights API documentation says its default reporting window is 7 days after a click and 1 day after a view. Check that each ad set still uses the setting you compare against, because changing it moves ROAS with no change in sales. Reporting tools are a second source of change. Meta's developer changelog says that from 12 January 2026 the Ads Insights API no longer returns data for 7-day and 28-day view-through windows: "attribution windows will no longer be available and will return no data". A reporting tool that built ROAS on those windows lost that credit on that date, which can look like a drop in a month when nothing else changed. Pass: your ad sets and tools used one setting across both periods. Fail: a setting or a tool changed on the date of the drop, so rebuild both periods on one window before you conclude anything.
Is it creative fatigue?
Meta's documentation defines frequency as "The average number of times each person saw your ad. This metric is estimated." The sign to look for is frequency rising while click-through rate falls on the same ad. Put frequency, click-through rate and cost per thousand impressions side by side for your top ads over the last several weeks. Pass: click-through rate holds as frequency rises, so the creative isn't tiring. Fail: click-through rate falls as frequency rises while price holds, so people are seeing the same ad more often and responding less. No page read for this post gives a frequency at which ads stop working, so read your own trend rather than a benchmark. Your Meta creative test is not an A/B test explains why a new ad can reach a different audience than the old one.
Did the auction get dearer?
Split ROAS into four numbers and compare this period with an equal period before: cost per thousand impressions (CPM), click-through rate, conversion rate and order value. ROAS is revenue per impression divided by cost per impression. So CPM up with the other three flat means price, CPM flat with conversion rate down means the site or the audience, and order value down means mix or discounting. Meta's 12% is a company-wide average as Meta reports it, not a price for your account, market or placement, and the release gives no reason for it, so your own CPM trend is the evidence that counts.
Is it the calendar or the tracking?
Compare the same weeks last year in Shopify before you blame the account: sessions, conversion rate and order value each move with the calendar and with demand, and falling consumer confidence is one example of a demand change. For tracking, count the GA4 sessions your Meta ads sent, tagged with utm_source and filed under Paid Social, which Google defines as arrival "via ads on social sites like Facebook and Twitter", and compare them with Shopify orders for the same days. Pass: Meta's purchases, those sessions and your orders fell together, so the drop is real sales, not counting. Fail: Meta's purchases fell while sessions and orders held, which points to a counting or tracking change, not a sales change.
Can Ads Manager tell you the drop is real?
Not by itself. Whether ads caused the orders they're credited with is a causal question, and Ads Manager counts purchases that Meta's ads touched without a comparison group. Gordon, Zettelmeyer, Bhargava and Chapsky (Marketing Science, 2019; two of the authors worked at Facebook) used data from 15 U.S. advertising experiments at Facebook, comprising 500 million user-experiment observations and 1.6 billion ad impressions, to compare the experiments' results with observational models. They report that "the observational methods often fail to produce the same effects as the randomized experiments". That is a peer-reviewed finding about observational methods on Facebook's own experiments, not a measurement of Meta's attribution setting. A holdout gives you the comparison group a dashboard lacks: Cut a channel? Run a holdout first.
Sources, 30 September 2026: Meta Reports Second Quarter 2026 Results (Meta, 29 July 2026); Ad Account Insights reference (Meta for Developers, 2026); 2025 out-of-cycle changes (Meta for Developers, 2025); A comparison of approaches to advertising measurement: evidence from big field experiments at Facebook (Marketing Science, 2019); Default channel group (Google, 2026)
Related answers
Frequently asked questions
Why did my Meta Ads ROAS drop this month?
Check in order: whether your attribution setting or reporting tool changed, whether frequency rose while click-through rate fell, whether your CPM rose, whether the same weeks last year dipped, and whether Meta's purchases diverged from GA4 sessions and Shopify orders.Have Meta ads got more expensive?
Meta's second-quarter 2026 release says average price per ad rose 12% year over year, as a company-wide average. That isn't your price. Compare your own CPM with the same weeks last year, and look at click-through and conversion rate beside it.What is Meta's default attribution setting?
Meta's Insights API documentation says the default reports actions within 7 days of a click and 1 day of a view. Check the setting on each ad set before you compare periods.
Go deeper: Causal attribution, explained.
Sixty-second versions of these ideas: Causality Engine on YouTube Shorts.
Have an audience of ecommerce operators? Earn 20% recurring commission by recommending the €99 read.
Keep reading
Terms in this article
- Ad ImpressionAd Impression is a single instance of an advertisement displaying on a webpage. Impressions are a key input for models measuring the causal impact of ad exposure on user behavior.
- AttributionAttribution identifies user actions that contribute to a desired outcome and assigns value to each. It reveals which marketing touchpoints drive conversions.
- Attribution SoftwareAttribution Software measures campaign impact by tracking customer interactions across touchpoints. It assigns value to each channel, showing what drives conversions.
- Attribution WindowAttribution Window is the defined period after a user interacts with a marketing touchpoint, during which a conversion can be credited to that ad. It sets the timeframe for assigning conversion credit.
- Conversion rateConversion Rate is the percentage of website visitors who complete a desired action out of the total number of visitors.
- ExperimentsExperiments are scientific procedures that test hypotheses or demonstrate facts. In marketing, experiments like A/B tests determine the causal effect of campaign changes, enabling data-driven decisions.
- IncrementalityIncrementality is the part of an outcome, such as orders or revenue, that a marketing activity caused: the extra results compared with what would have happened without it. It is a causal effect, not the number of conversions credited to a channel.
- Incrementality TestingIncrementality Testing measures the additional impact of a marketing campaign. It compares exposed and control groups to determine causal effect.