Skip to main content
A targeting experiment is an audience-level group that sits beneath an initiative. Each targeting experiment defines who sees your ads — their age, gender, location, interests, and daily budget. Because each initiative can have multiple targeting experiments, you can A/B test completely different audiences within the same campaign without duplicating your creative work.

​
Add a targeting experiment

1

Select the initiative

On the funnel canvas, locate the initiative you want to add targeting to.
2

Add a targeting experiment

Click the + button on the initiative node. A new targeting experiment node is created and connected below the initiative.
3

Name the experiment

Click the targeting node to open its settings panel. Enter a name that describes the audience — for example, “Women 25–44 US broad” or “Lookalike 1% purchase list”.
4

Configure targeting parameters

Set the audience parameters in the settings panel:
  • Targeting type — Choose the audience strategy:
    • broad — Let Meta’s algorithm find the best audience within your other constraints.
    • interest_stack — Target by specific Meta interests. Add interests by name; Fig resolves them to Meta interest IDs.
    • lookalike — Target users similar to a source audience in your Meta ad account.
    • retargeting — Reach people who have previously interacted with your ads or pages.
  • Age range — Set the minimum (default 18) and maximum (default 65) age.
  • Gender — Target all, male, or female.
  • Geo locations — Add one or more countries, regions, cities, or radius targets. Defaults to the United States if left empty.
  • Interests — If using interest_stack, search for and add Meta interest targets. Fig stores the canonical Meta interest IDs so your targeting is accurate when published.
  • Conversion event — Specify the Meta pixel event to optimize for (e.g. Lead or Purchase).
  • Daily budget — Set a budget at the targeting level. Use this when the parent initiative has its budget type set to set at group.

​
Multiple targeting experiments per initiative

Add as many targeting experiments to an initiative as needed. Each experiment appears as a separate node below the initiative on the canvas, and each can have its own unique audience settings and daily budget.
Running two targeting experiments simultaneously — for example, broad vs. interest-stacked — is an effective way to discover which audience delivers better results for the same offer and creative.

​
Targeting statuses

Targeting experiments share the same status lifecycle as initiatives:
Moving a targeting experiment to syncing, review, or live requires its parent initiative to be properly configured with a connected Meta account.

​
Reorder targeting experiments

Use the left and right arrow controls on a targeting node to change its position under the parent initiative. This is a visual change only.

​
Delete a targeting experiment

Open the targeting node’s settings panel and select Delete. Deleting a targeting experiment removes it and all the ad assets beneath it from the workflow.