Graduate or Scale Where It Lies
Choose between isolating a proven concept and scaling it in place.
- Difficulty
- Moderate
- Time to result
- ~weeks to results
- Steps
- 5
- Confidence
- 95%
Graduate or Scale Where It Lies is a two-path rule for expanding a proven ad. If the ad is performing well in its current campaign and has room to spend, leave the structure intact and increase the budget where it already works. If the team wants to develop a validated concept more deliberately, graduate it: duplicate the ad into another campaign, isolate it, and create purposeful variations around the known winner. A successful video might produce a shorter clip, a still image, a motion graphic, and a retained original. The distinction prevents unnecessary restructuring while giving a strong concept dedicated space when that structure serves a clear test. In both paths, the aim is to preserve the signal that already works rather than adding many unrelated ads that fragment budget and learning.
Origin
Extracted from The Foundr Podcast
Core principles
- 01Do not disturb an ad that is already scaling effectively.
- 02Isolate a proven concept when focused variations need room to develop.
- 03Derive new formats from the element that has already demonstrated demand.
- 04Campaign structure should express a hypothesis, not merely contain more ads.
How to run it
- 1
Locate the real winner
Identify which ad and underlying concept are producing the useful outcome, rather than relying on account-level averages.
Pro tip Separate the concept from incidental details such as room, crop, or minor edit.
Watch out A low-spend ad has not necessarily earned winner status.
- 2
Assess room in the current setup
Check whether the current campaign is working cleanly and can take more budget without requiring structural change.
Pro tip Prefer the in-place path when the existing setup is already allocating effectively.
Watch out Duplicating by habit can split spend and reset learning without adding insight.
- 3
Scale where it lies
When performance is healthy, increase the budget in the current campaign and monitor the response.
Pro tip Apply the rule: do not fix what is not broken.
Watch out Increasing budget still requires trend monitoring and known unit economics.
- 4
Graduate when isolation has a purpose
Duplicate the proven ad into a separate campaign when focused development of the concept needs dedicated budget and measurement.
Pro tip Write the hypothesis for isolation before duplicating anything.
Watch out Isolation without a hypothesis is just extra complexity.
- 5
Build concept-led variations
Turn the winning video or image into related clips, stills, motion graphics, or other formats while preserving the reason it worked.
Pro tip Keep the original running as a control.
Watch out Minor cosmetic edits may create more ads without producing a meaningfully different signal.
In the wild
Shackleford describes a video of a woman presenting a product. Under the graduation path, the team duplicates and isolates the proven ad, then derives a clip, a still image, and a motion graphic while keeping the original video. Under the alternative path, if the original campaign is already working, the team simply increases its budget in place.
→ The team chooses structure based on the scaling purpose instead of duplicating every winner automatically.
Common mistakes
Graduating every winner automatically
Duplication can add complexity and fragment spend when the original campaign already has room to scale.
Calling cosmetic edits diversification
A new room or minor visual edit may still present Meta and the customer with the same underlying signal.
Removing the original control
Turning off the proven ad makes it harder to judge whether derived formats add value.
Is it for you?
Best for
Paid-social teams with a proven ad that must decide how to pursue additional scale.
Not ideal for
Accounts without a clear winning concept or enough budget to support an isolated campaign.
From the transcript
“I do believe in a graduation period or what we call scale it where it lies.”
“Rule one, don't fix it if it's not broken. Increase the budget.”
From the episode
635: The Meta Ads System Working in 2026