Measurement
Govern spend by contribution margin. Let lift and holdout tests, not Meta's reported ROAS, decide budget.
24 answers · 8 new or changed since Nov 16 · Data as of Nov 22, 2026

The short version
Our judgment, drawn from the 24 answers below.
Set a monthly contribution-margin target and derive your Meta ROAS target from it, instead of running to a fixed ROAS number.
Run Meta's free Conversion Lift test once you qualify, and scale one step only when the low end of a test's result clears breakeven.
From How often to run incrementality tests, Acting on a finished lift or geo test
Use store orders and revenue as the number of record, and never add platform-reported conversions together.
From Reconciling Meta with GA4 and store totals, Comparing results across 2026 reporting changes
Run the Conversions API alongside the Meta pixel with a shared event ID, and keep one pixel per business in your own portfolio.
From Auditing your tracking setup, One Meta pixel or several
The 24 answers
Grouped by the decision you are making. Jump to a group:
Pick the numbers that count
Which business numbers govern spend, how to reconcile reports and how to read a weak week.
Which business metric should govern our Meta spend?
Govern Meta spend by a monthly contribution-margin target. Derive your total sales divided by total ad spend (MER) and Meta ROAS targets from it, instead of running to a fixed ROAS or MER number.
Medium confidence · 16 sources
How do we reconcile Meta's numbers with GA4 and store totals?
Use store orders and revenue as the number of record. Use Meta's conversions only to compare campaigns inside Meta, after an incrementality adjustment (see “How much Meta ROAS overstates return”), and lift tests for what Meta caused. Never add platform-reported conversions together.
Updated Nov 16 Medium confidence · 16 sources
How do we compare Meta results across the 2026 reporting changes?
Treat January 12, 2026, March 2026 and November 12, 2026 as reporting breaks. Compare years and before-and-after periods on backend revenue, orders or new customers, not Meta-reported conversions, and judge Black Friday on store orders.
Medium confidence · 13 sources
How do we tie repeat purchases back to the ad that first brought each customer in?
Tag every customer's first order with the campaign and ad that brought them in, and keep that tag on every later order. Then compare cohorts at the same age on repeat rate and contribution, not first-order CAC.
Low confidence · 10 sources
When our Meta cost per purchase or ROAS gets worse week over week, how do we find which lever moved it: CPM, click-through rate, conversion rate or order value?
First confirm that store revenue and orders fell too. Then put CPM, click-through rate, conversion rate and order value side by side, and fix the lever that moved: auction, ad, site or product mix.
Medium confidence · 7 sources
What click-through rate and cost per click should our Meta ads get, and when is ours too low?
Expect a link click-through rate near 1.64%, the median across MHI Growth Engine's consumer brands, and rework the ad below 1.23%. Those brands paid a median $1.66 a link click. A costly click with a normal rate means costly impressions.
New Nov 17 Medium confidence · 8 sources
Fix the tracking
Pixel and Conversions API setup, consent settings and app event tracking.
Is our tracking setup complete enough (Conversions API, deduplication, match quality)?
Run the Conversions API alongside the Meta pixel with a shared event ID. Keep purchase event match quality above 5 (Common Thread Collective aims for 7). At larger scale, add a dedicated server-side tool and a specialist monitor.
Medium confidence · 13 sources
Should we use one Meta pixel or several across our sites and ad accounts, and who should own it?
Run one Meta pixel, which Events Manager calls a dataset, per business, owned in your own business portfolio and shared with agencies. Add a second only for a separate business or legal entity. Don't count on deleting or moving one.
New Nov 17 Medium confidence · 8 sources
Why does our Meta pixel show as not active or not working, and how do we fix it?
Inactive means the pixel isn't sending events correctly. Check your site with Meta Ads Data Advisor and Test Events, then fix the first break: missing code, a wrong pixel ID, consent or ad blockers, or a paused Shopify app pixel.
New Nov 17 Medium confidence · 5 sources
Should we leave Meta's on-device conversion setting on?
Keep the setting on where your consent setup is documented and you track Meta purchases against store orders from before November 12. Switch it off in health, finance or other sensitive businesses, or wherever consent handling is unclear.
Medium confidence · 12 sources
Which attribution setup should our iOS app campaigns on Meta use?
Use Meta's Aggregated Event Measurement (AEM) when your app and event are eligible, and send each event once. Keep SKAdNetwork as a separate view across networks. Judge both against App Store Connect and your revenue backend.
Medium confidence · 5 sources
Read Meta's attribution
Which attribution setting to use, how far to trust it and what to check it against.
Which attribution window and setting should each campaign use?
On campaigns not yet on Incremental Attribution, keep Meta's default setting of 7-day click, 1-day engage-through and 1-day view. Judge results on 7-day click alone.
Medium confidence · 15 sources
Should we switch to Meta's Incremental Attribution setting?
Expect about parity with standard attribution. Switch to Incremental Attribution only if your own test shows a gain.
Updated Nov 18 Low confidence · 4 sources
How much does Meta-reported ROAS overstate our real return?
Until your own lift test, read acquisition campaigns' 7-day click ROAS as reported, discount retargeting and existing-customer ROAS, and leave 1-day view out. The benchmark factors mostly predate Meta's March 2026 click change, so test soon.
Updated Nov 19 Medium confidence · 9 sources
Should Meta optimize on our third-party attribution tool's results?
Try Custom Attribution only through a two-cell test in new campaigns, against the Meta setting you would otherwise run. Do it only if you pay for a supported tool whose numbers differ from Meta's. Otherwise stay on Meta's models.
Low confidence · 9 sources
Do we need marketing mix modeling or a paid multi-touch attribution tool?
Depends on your spend. Open it to see the answer for each spend range.
Medium confidence · 12 sources
Should we run a post-purchase survey, and how much should it count against Meta's numbers?
Run a one-question post-purchase survey and read it beside Meta's reporting as a second check. Give it an answer for every channel you pay for. Let lift tests, not the survey, decide Meta's budget.
Medium confidence · 7 sources
Run lift tests
How to design, run and act on lift, geo and setting tests, and feed them into a mix model.
Should we run incrementality tests, and how often at our spend?
Run Meta's free Conversion Lift test once you qualify, starting at the account level. Retest Meta after any big change in budget, media mix or attribution setting.
Medium confidence · 12 sources
How do we design a lift or geo test for our order volume?
Set the pass bar at your breakeven lift before launch, then size the test to detect it. For a geo test, hold out at least 15% of markets, run four weeks or more and read once at the end.
Updated Nov 17 Medium confidence · 16 sources
How do we turn a finished lift or geo test into a budget move?
Read a finished test's interval against breakeven. Scale one step (about 15 to 20% in Stella's examples) when its low end clears it, hold and retest when it straddles and cut in steps below. Move targets only partway.
Medium confidence · 11 sources
How do we test a new Meta setting against our current setup?
Test a new Meta setting as two matched cells at equal budget, about 100 purchases each over at least two weeks, ignoring week one. Decide on new-customer cost or margin in your own numbers, not Meta's.
Medium confidence · 13 sources
How much of Meta's impact lands off our site, and how do we credit it?
If you sell only on your own site, judge Meta on total site revenue in geo holdouts or lift tests. These capture its effect on branded search and direct traffic.
Updated Nov 19 Medium confidence · 4 sources
Should we measure brand lift or search lift from Meta, and how?
Measure search lift by adding search visits to the Meta lift test you already run. Run Brand Lift only for an awareness flight once you meet Meta's $120,000 minimum. Let sales tests decide budget.
Medium confidence · 9 sources
How do we calibrate a mix model with our lift and geo tests?
Feed each finished lift or geo test into your mix model as a prior that keeps its interval. Weight several tests by relevance, recency, precision and spend. Refresh on your budget cycle, and recalibrate early after big changes.
Medium confidence · 15 sources
More in Meta
186 answers in 10 subjects
Guides that use these answers: How to tell if your ad spend is working · How to market a lead-gen business on Meta and Google · How to start advertising on Meta