The clearest way to see what Custom Attribution does for you is a two-cell A/B test: one campaign optimizing against your Triple Whale attribution, one business-as-usual campaign on Meta's standard attribution, compared in Triple Whale.
Before you begin
Attribution Passback is enabled and your connection quality shows High — Set up Meta Custom Attribution.
You know how to configure an eligible campaign — Build a Meta campaign with Custom Attribution.
You can run the test for at least 2 weeks. It can run in any region.
You set the test up yourself. Meta does not configure the campaigns or the experiment for you. Loop in your Meta contact if something has to change mid-flight, but the build is yours.
Test design
Two cells, identical in everything except the attribution model:
Cell | Setup |
Test | Meta PBA test – TEST {campaign name} — attribution model set to Custom |
BAU (control) | Meta PBA test – BAU {campaign name} — attribution model set to Standard |
Creatives and targeting must match across both campaigns.
Once the test is live, make no manual changes to either campaign for the full duration.
Step 1: Build your Test and BAU campaigns
In Meta Ads Manager, build two campaigns. The quickest route is to duplicate an existing reference ad set into each so the setups are identical.
Control - Create a new Sales objective campaign and, within the new campaign, create a duplicate (ad set) from the reference (ad set). Name this “Meta PBA test – BAU {campaign name}”.
Test - Repeat step 1 and name this “Meta PBA test – TEST {campaign name}”
Campaign and ad set need to satisfy the following eligibility criteria (for both Test and Control)
Select Sales as your campaign objective, then click Continue.
At the campaign level, select Campaign budget as your budget strategy, and Highest volume or Highest value as your Campaign bid strategy.
Create and edit an ad set to configure these conversion settings:
Conversion location: Select Website or Website + App.
Performance goal: Select Maximize number of conversions or Maximize value of conversions
Dataset: Select an eligible dataset
Conversion event: Select Purchase
Attribution model:
Select Custom for Test
Select Standard for BAU
Select the custom attribution source:
If using Web only, custom attribution source will be auto-populated.
If using both Web and App datasets, select the custom attribution source for each.
Save and stop for QA
Reminder: Please make sure creatives and targeting match for both test and control.
BAU (control) Campaign Example
Duplicating an existing reference ad set into a new Sales campaign named "Meta PBA test – BAU", setting Campaign budget and the Highest volume bid strategy, and confirming the attribution model is set to Standard. The campaign is left paused for QA.
Don't see a Custom option for the Test Campaign?
If the Attribution model dropdown shows only Standard and Incremental, Custom Attribution is not yet provisioned on your ad account. Meta hides the option entirely rather than disabling it, so there is no error message to tell you what's wrong.
Check two things:
Attribution Passback is enabled in Triple Whale — see Set up Meta Custom Attribution.
Your connection quality shows High in Meta Events Manager. Meta needs roughly 14 days of high-quality signal before the option appears — see Confirm Custom Attribution is available.
Until both are true, the Custom option will not appear and you cannot build the Test cell. Nothing is misconfigured in your campaign.
Step 2: A/B Test Setup
Use Meta's Experiments tool to split your audience between the two cells.
Open Experiments and click Create an A/B test.
Add your campaigns to the two cells:
Campaign A (Test):
Meta PBA test – TEST {campaign name}Campaign B (BAU):
Meta PBA test – BAU {campaign name}For more complex designs, such as a 3-cell test, add additional cells and campaigns as needed.
Set the schedule:
Start one day BEFORE your campaign start date
End one day AFTER your campaign end date
DISABLE ‘End test early if a winner is found’ to preserve statistical validity
Test Name
Name your study clearly (e.g., Custom Attribution vs BAU - Q3 2026)
Determining the winner:
Select ‘Cost per purchase’ as the key metric for winner declaration
Advertisers may add metrics for additional business-specific KPIs
Advertisers should use their own Source of Truth to determine campaign winner
Step 3: Launch and monitor
Things to monitor during the test flight
Continue to monitor the following key metrics and reach out to Meta with any questions:
Ensure CAPI integration & campaign quality for changes in Events Manager/Ads manager
❗Reminder: No manual changes are to be made to test/control campaigns during the duration of the test. If any changes are made please alert Meta immediately.
Step 4: Review results in Triple Whale
Custom Attribution campaigns are tagged with a "Custom" label in Ads Manager. Rather than displaying Meta's default attributed conversions, the Results column reflects the advertiser’s externally attributed data.
We strongly recommend advertisers evaluate the performance of the Custom Attribution campaign against their BAU campaign in Triple Whale. This is due to 2 factors:
Difference in attribution methodology. In Ads Manager, Custom Attribution results may show lower KPIs compared to Standard Attribution campaigns. This is because Custom Attribution reflects cross-publisher attribution, whereas Meta's standard reporting uses single-source attributed results.
Data processing timing. There may be data processing delays/differences in attributed conversion data Meta receives from the external analytics tool.
Scaling after the test
If the Test cell wins in Triple Whale, extend Custom Attribution beyond the test campaign. Two approaches:
Account-level split. Run a 50/50 split between Custom Attribution and standard campaigns to keep measuring as you scale.
Budget floor. Move at least 20% of your Meta budget onto Custom Attribution campaigns. Below that, the signal is too thin to read.
FAQ
How long does the test need to run?
At least two weeks, on top of the roughly 14 days of signal collection needed before Custom Attribution becomes available.
Can I end the test early if one cell is clearly winning?
No — and disable Meta's option to do it automatically. Calling a test on partial data invalidates the comparison.
Does Meta set the test up for me?
No. You build the campaigns and configure the experiment. Your Meta contact can advise, and should be told immediately if anything changes mid-flight.
Is there a minimum spend?
Meta does not publish a spend minimum for A/B tests in the Experiments tool. What matters more is conversion volume — a test needs enough purchases in each cell to produce a readable difference. If your Meta spend is modest, talk to your Meta contact before committing to a two-week flight.