Monthly Growth Lever Loop
Isolate one growth lever, test it, double down, then replace it at saturation
- Difficulty
- Moderate
- Time to result
- ~months to results
- Steps
- 5
- Confidence
- 96%
Yoon describes a monthly discipline built around the mantra that the company had to get smarter every month. She used a spreadsheet to name growth levers, record the variables present during each period, and try to isolate their effects through extensive A/B testing. When a lever showed a clear contribution, the company increased its focus on it. When the lever eventually reached saturation and stopped scaling efficiently, the search moved to the next one. The mechanism is not perfect attribution. Yoon acknowledges that top-of-funnel activity is hard to measure and variables can be conflated. Instead, the framework creates cumulative learning under uncertainty: log the context, design the cleanest practical comparison, act on strong evidence, and keep testing the lever's limits rather than assuming it will work forever.
Origin
The loop emerged while Yoon was bootstrapping Peach and Lily with very little cash. The need to avoid vague marketing judgments pushed her toward spreadsheets, variable tracking, and repeated testing.
Core principles
- 01Get smarter about growth every month
- 02Record the variables surrounding each result
- 03Isolate causes through controlled tests where possible
- 04Double down only when a lever clearly works
- 05Expect every efficient lever to reach a limit
How to run it
- 1
Name the lever
State the specific marketing action you believe can drive growth. Avoid grouping several unrelated tactics under one label.
Pro tip Phrase the lever so another person could repeat the same action.
Watch out A vague label makes later attribution meaningless.
- 2
Map the variables
Record other campaigns, pricing changes, seasonality, and operational factors active during the test. This makes possible confounders visible.
Pro tip Keep a recurring spreadsheet with the same core fields each month.
Watch out Unrecorded simultaneous changes can create false confidence.
- 3
Run the cleanest comparison
Use an A/B test or another bounded comparison when practical. Define the outcome and observation period before judging the result.
Pro tip Prefer tests that change one major variable at a time.
Watch out Not every top-of-funnel effect will be directly attributable.
- 4
Double down on evidence
Increase attention or spend when the lever produces a clear, repeatable improvement. Continue measuring as scale changes.
Pro tip Scale in increments so a reversal remains affordable.
Watch out One strong result may be noise rather than a repeatable lever.
- 5
Detect saturation
Watch for diminishing efficiency as the lever grows. When it taps out, preserve what still works and begin testing the next candidate.
Pro tip Compare marginal results, not only total results.
Watch out A formerly efficient lever is not automatically scalable forever.
In the wild
A small store records one email campaign as the candidate lever, notes a simultaneous promotion, and compares conversion with a control group. After the lift repeats, it expands the campaign gradually. When marginal returns later flatten, the team keeps the profitable core and tests a different acquisition channel.
→ The team compounds learning without betting the whole budget on a single result.
Common mistakes
Calling a feeling a result
Saying a tactic generally worked without isolating variables does not create reusable learning.
Scaling past efficiency
A lever that works at small scale may stop producing acceptable marginal returns.
Is it for you?
Best for
It is best for teams that can run bounded marketing tests and review results on a recurring cadence.
Not ideal for
It is not ideal when the measurement window is too short to observe the intended result.
From the transcript
“you just have to get smarter every month and understanding what marketing tactics 28 30 actually work or not”
“When you see something clearly working like then double down on that.”
From the episode
581: How to Build a $100M Brand Without Raising a Dollar