What Should You Do With ROAS Targets When Meta's Reported Conversions Jump?

Do not change a ROAS target just because Meta reports more conversions. From November 12, 2026, Meta's “Measure conversions on customer devices” setting is scheduled to switch on by default, so reported conversions can rise without more sales. Set a baseline from store revenue before then and judge targets against it.

Targets are the part of a paid account that moves money. If Meta's report improves overnight, the pull to raise budgets or loosen a target is strong, especially in Cyber Week. This post covers what to hold steady, what to compare, and what is still unknown. For the full explanation of the setting itself, start with why Meta can report more conversions than your store has orders.

What do we actually know about the change?

According to industry coverage, Meta is turning on a Pixel setting called “Measure conversions on customer devices” from November 12, 2026, on by default for eligible advertisers. It matches some conversions to ad interactions on the shopper’s device instead of only on Meta’s servers. Coverage citing Meta’s help pages says it does not change the configured attribution window or attribution model, and that it is not active for people in the EU and UK. You can turn it off per dataset in Events Manager. Confirm the status of your own Pixel there.

Could reported ROAS go up without better performance?

Yes, that is possible. One source says advertisers may see more attributed conversions and that reported CPA could fall and ROAS could rise even if sales and spend stay the same, and notes that an increase in attributed conversions does not necessarily mean more actual conversions. Another points out that Meta has not said whether purchase value and currency are sent with the recovered conversions, which decides whether they show up in ROAS at all. So the honest answer is that a ROAS change is possible and unconfirmed. A new number in the report is a reason to check, not a reason to act.

Does it change how Meta bids?

Meta has not said. Coverage notes that if Meta uses the recovered conversions for optimization, delivery could get a different signal, but that is unconfirmed. Do not explain a shift in delivery or cost by this setting alone, and do not claim it either way. Note when it happened and keep watching.

What should you do with ROAS targets and budgets?

Start with a baseline. Before November 12, record store revenue and Meta-reported revenue for each campaign over the last two to four weeks. That is the reference every later number gets compared with, and it is the same step practitioners recommend ahead of the change.

Then avoid mechanical moves in the first days. If Meta’s reported ROAS jumps, do not tighten or loosen targets purely on that jump. If you must adjust during Cyber Week, change in small steps, and base it on store revenue and margin. The standing rules in platform ROAS versus store data and how to improve ROAS still apply: the platform number is one view, your store is the other.

Finally, compare blended results as well as by campaign. Total store revenue against total ad spend does not depend on how any one platform counts, so it is the steadiest read through a measurement change. A widening gap between Meta’s revenue and store revenue after November 12 is the signal that counting changed, not demand.

What if store results do not follow the report?

If clicks and reported conversions look healthy but orders do not, check the store before the account. Broken steps and slow pages waste paid traffic however the platform counts it, as finding site issues that waste ad spend explains, and a Cyber Week checklist like the hour-by-hour monitoring runbook keeps the store side covered. If you want a free first look, run the free website audit.

How does Noibu fit in?

The ROAS agent follows paid traffic from the visit through to the order on your own site, so you can see what shoppers did after the click rather than relying on one platform’s count. It works in the Noibu console and through Noibu MCP in AI tools like Claude, and nothing ships until you approve it. Noibu is AI for ecommerce operations, the always-on optimization system that keeps online stores running at their best. A team of supervised AI agents finds what’s broken, slow, or leaking revenue, does the work, and hands results back for human approval. Trusted by retailers like Free People, Guess, Liquid I.V., and Ariat.

Frequently asked questions