Creative Smoke Test
Earn each increase in creative investment with evidence from the last stage.
- Difficulty
- Easy
- Time to result
- ~weeks to results
- Steps
- 5
- Confidence
- 99%
The Creative Smoke Test is a staged investment ladder for paid-social concepts. Begin with the product, setting, and bold message expressed in a simple static image, created from an existing shoot or with AI. Use early performance signals such as CPM, clicks, and carts to ask whether there is enough smoke to justify another stage. Convert a promising image into a basic motion graphic, then into employee-generated or user-generated content. Only after the angle continues to perform should the team pay for full video production and create many assets around it. Each stage raises cost and production effort while preserving the validated concept. The method reduces speculative creative spend and avoids mistaking a brief performance spike for a durable angle.
Origin
Extracted from The Foundr Podcast
Core principles
- 01Validate the concept before funding expensive production.
- 02Increase fidelity and cost only after the prior stage shows promise.
- 03Test the same angle across formats to broaden placement evidence.
- 04Sustained performance matters more than one strong weekend.
How to run it
- 1
State the concept
Choose the product, usage setting, and message you want to test. Keep the idea consistent across the first variations.
Pro tip Use a bold message that makes the proposed angle unmistakable.
Watch out Testing multiple propositions at once hides which concept earned the response.
- 2
Launch static smoke tests
Create inexpensive still images in a few relevant settings and run them as ads. Look for directional evidence in reach, clicks, and cart behavior.
Pro tip Existing product photography or a clearly disclosed AI-assisted scene can keep this stage fast and inexpensive.
Watch out Early signals are evidence to continue testing, not proof of profitable scale.
- 3
Add simple motion
Turn the strongest static concept into a motion graphic, such as a moving product or a short action. This raises production investment and reaches additional placements.
Pro tip Preserve the validated message while changing the format.
Watch out A format change that also replaces the concept is a new test.
- 4
Add human delivery
Develop the surviving concept as employee-generated or user-generated content before committing to a major shoot.
Pro tip Use people who can represent the product truthfully and naturally.
Watch out Do not fabricate testimonials, identities, or product outcomes with AI.
- 5
Fund full production
Invest in polished video only after the concept has performed across earlier stages and over enough time to reduce the chance of a temporary spike.
Pro tip Produce multiple assets around the proven angle once confidence is earned.
Watch out One exceptional weekend is not repeated validation.
In the wild
Shackleford proposes static images of a beverage at a bar, campfire, and home while keeping the core replacement message consistent. A promising image graduates to a spinning can or a person picking it up, then to employee or user-generated content, and finally to a full video investment after the angle continues to perform.
→ Production spending rises only as the concept earns stronger evidence across formats and time.
Common mistakes
Starting with the expensive shoot
A polished asset cannot rescue a concept that has not demonstrated customer interest.
Treating weak signals as certainty
CPM, clicks, and carts can justify the next test, but they do not independently establish profitable demand.
Using deceptive AI people
Synthetic spokespeople, testimonials, or transformations can misrepresent who endorsed the product and what it does.
Is it for you?
Best for
Paid-social teams that need a repeatable way to allocate creative production budget under uncertainty.
Not ideal for
Compliance-sensitive campaigns where even low-fidelity test assets require full legal or medical review before exposure.
From the transcript
“I believe in the smoke test.”
“The image works and performs.”
“Performance looked good at the smoke.”
From the episode
635: The Meta Ads System Working in 2026