Question

Which campaign structure is best for 2025 for TOF testing?

Option 1 (Multiple products in 1 ad set, targeting different angles):

Campaign 1

• Ad Set 1 = Angle 1 (problem) = Multiple Products (3-4 ads, different formats)

• Ad Set 2 = Angle 2 (problem) = Multiple Products (3-4 ads, different formats)

Option 2 (One product per ad set with different angles for multiple products):

Campaign 1

• Ad Set 1 = Angle 1 (problem) = Product 1 (3-4 ads, different formats)

• Ad Set 2 = Angle 2 (problem) = Product 2 (3-4 ads, different formats)

Option 3 (Separate product per campaign):

Campaign 1

• Ad Set 1 = Angle 1 (problem) = Product 1 (3-4 ads, different formats)

Campaign 2

• Ad Set 1 = Angle 2 (problem) = Product 2 (3-4 ads, different formats)

Answer 1

Based on both Meta’s current algorithm behavior and community insights, Option 2—one product per ad set with different angles—is the most balanced and effective campaign structure for TOF testing in 2025. 

It allows for cleaner audience signals, better attribution, and easier optimization without overwhelming the algorithm. This setup works especially well if you're testing multiple creative angles per product and want to scale what performs best.

Option 1, which combines multiple products in one ad set, can work too—particularly for broad product categories where angle testing is the priority. However, it may mix signals and slow down optimization.

Option 3 (separate campaigns per product) offers the most control but is better suited for higher budgets, as it splits budget across multiple campaigns. For most advertisers looking to test efficiently, Option 2 strikes the right balance between clarity, scale, and performance.