← Back to Playground
Playground MMM simulator

Interactive MMM workbench

Adstock & Saturation Simulator

Adstock is not about making media look bigger. It is about matching the timing of media exposure to the timing of consumer response.

Open simulator 20 weeks · client-side · no backend

Active adstock function

Geometric (recursive)

Tracks the selected adstock function and uses MMM article notation: media input M, adstocked media A.

Current media plus a fraction λ of last week’s adstocked media. Memory accumulates — the default MMM choice for channels with an immediate peak.

Controls

Shape the transform

Fine-tune how media exposure becomes effective media, then how effective media bends into response.

Adstock function

Adstock Shape

Compare geometric carryover functions across λ values.

λ = 0.2
λ = 0.5
λ = 0.7
λ = 0.9
Your current λ = 0.70

Each curve starts at 100% because recursive geometric adstock is an unnormalized memory reservoir. Higher λ means slower decay and a larger long-run carryover multiplier. The blue curve tracks your current λ; the dot marks its value at lag 3.

Saturation Shape

Vary α to see how steepness changes with θ fixed.

α = 0.5
α = 1
α = 2
α = 5
Your current α = 1.7

θ is fixed at 280, so every curve crosses 50% response at the same spend level. Higher α makes the bend sharper around θ. The blue curve and dot mark your current K and α — K always sits at the 50% response line.

Raw Media Spend

Actual media spend by week before any transformation.

Actual media spend

Effective Media After Adstock

Effective media, not actual spend. The dashed line keeps raw spend visible as a reference.

Effective media
Actual spend reference

Saturated Response

Hill saturation compresses high effective-media weeks into diminishing returns.

Saturated response