Susan McBride › MadCat Consulting › Meta Creative Planner

CreativeMath

How many new concepts vs. variations should you test?

Turn your Meta media budget into a practical monthly creative plan. Balance proven winners, fresh concepts, creative variations, and meaningful testing capacity.

1. Set your assumptions

These are planning inputs, not universal benchmarks. Adjust them to reflect your brand, category, creative resources, and level of competition.

$5,000–$5,000,000; logarithmic scale gives more control at lower spend.
Cost to acquire one customer or achieve your selected conversion goal.
The rest is planned for proven winners and scaling.
Spend allowance per execution, expressed as a multiple of target CPA. 15× CPA implies ~15 conversions only if target CPA is achieved.
Sets a suggested share of screening slots for new concept executions. You can override it.
Remaining screening slots go to variations of existing winners.
Different creators, hooks or formats used to bring one new strategic idea to life.

2. Your monthly media allocation

Separate the budget intended to scale proven ads from the budget reserved to learn from new creative.

Proven winners / scaling
Creative testing
These are planning allocations, not a promise that Meta will distribute spend evenly or that winners must live in separate campaigns. Newly successful ads can graduate into the scaling pool.

3. Your monthly creative plan

Five distinct parts of the program. Counts are planning estimates, not mandated production quotas.

1 · SCALE

Evergreen winners

Planned spend behind ads already delivering acceptable business results. This is a budget, not a recommended number of evergreen ads.
2 · IMPROVE

Variations of winners

New executions of proven ideas: a different creator, hook, edit, demonstration or format. These consume part of the testing budget until they prove themselves.
3 · EXPLORE

New creative concepts

Distinct strategic ideas, messages or customer insights—not just a new ad ID, landing page, or another creator reading the same script.
4 · EXECUTE

New concept executions

Actual ad assets made from the new concepts. For example, two creators or hooks expressing one new concept count as two executions, but only one concept.

5 · Meaningful testing capacity

Estimated number of individual creative executions that can each receive the chosen initial screening allowance. This is not statistical significance, guaranteed delivery, or a prediction of how many winners you'll find.

What counts as new creative?

New concept: A new proposition, consumer tension, benefit, proof point or narrative. Example: moving from a product demonstration to a story about a specific unmet customer need.

Variation of a winner: A fresh execution of an already proven concept. Example: a new creator, opening hook, edit or format expressing the same core message.

New concept execution: One finished ad asset bringing a new concept to life. Multiple executions can belong to the same concept.

Not counted: Landing page swaps, duplicate ad IDs, and cosmetic edits without a meaningful creative change. Landing page tests remain valuable but should be tracked separately.

Methodology: Testing budget ÷ (target CPA × screening multiplier) gives the maximum whole initial screening slots. Slots are divided between variations of winners and executions of new concepts. New concept count is derived from executions per concept. Any unused slots are shown rather than silently reallocated. Higher production volume is possible, but does not mean every asset can receive an adequately funded test. A statistically powered experiment requires a separate sample-size design.