A campaign that spikes and plateaus is a media buy.

A system that compounds is a business asset. The difference shows up in the growth curve. A media buy produces a peak followed by regression to baseline. A compounding system produces a floor that rises each month, because each month's results feed the next month's targeting.

The Protein Pals growth curve

Protein Pals is a high-protein Indian meal delivery service in the Toronto GTA. When Daee Media began working with them in May 2025, there were 15 active subscribers. By October 19th, when the kitchen hit capacity and the founders paused the campaign, the subscriber base was at 230-240.

15 Subscribers in May Before any paid campaigns
45 New in 15 days First lead gen campaign · Jul 2025
240 Active subscribers At campaign pause · Oct 2025

That is not a 16x growth story from one good campaign. It is a 16x growth story from a system that accumulated signal, improved targeting, lowered CPL, and converted leads at an increasing rate month over month.

The compounding mechanics

Month 1
June 2025: Foundation

Engagement ads running at $0.50 per conversation. ManyChat flows tested. Pinned carousels published. The warm audience pool building before lead gen launches. No subscriptions from paid yet, but the audience is not cold when the campaign starts.

Month 2
July 2025: First conversion cohort

45 new subscriptions in the first 15 days of lead gen. The first converter data flows back to Meta via CAPI. The algorithm gets its first real signal: this is what a Protein Pals subscriber looks like.

Month 3
August-September 2025: Crisis and hold

Meta rejection cycle. CPL climbs to 8-9 CAD. Volume drops. But the CAPI data from the first two months does not disappear. It waits in the model.

Month 4
October 2025: Compounding visible

Relaunch day: $1.28 CPL. 33 leads in 12 hours. The algorithm is running on four months of outcome data. It is not finding leads anymore. It is finding subscribers. The compounding is now visible in the number.

July 13, 2025 · 45 closed in 15 days: the first conversion cohort that fed the CAPI model
July 13, 2025 · 45 closed in 15 days: the first conversion cohort that fed the CAPI model

Why 420 leads in pipeline at pause is not a failure

When the campaign paused on October 19th, 420 leads were still in the pipeline: qualified but not yet called, or contacted but not yet closed. Most businesses would read this as a conversion failure.

It is not. It is a warm list that the next campaign does not need to generate from scratch. When the kitchen upgrades and campaigns restart, those 420 leads receive a re-engagement message. Some will convert in the new year when the routine motivation resets. Some will refer. The pipeline does not expire. It waits.

A system that compounds does not reset to zero when you pause it. The data stays. The audience model stays. The pipeline stays. You restart from a higher floor than you started from in June.

Read the full Protein Pals engagement: the system, the numbers, and the pause.

Read the case study →