What is saturation?
Saturation measures how much of a channel's realistically reachable effect current spend already captures — a value between 0 and 1 on the channel's fitted diminishing-returns curve. Near 0, the channel is barely tapped and additional spend buys nearly proportional margin. Near 1, the audience is exhausted: more budget buys impressions, not incremental sales. The model stores it as a decimal from 0 to 1; the app displays it as a percentage (0.6 shows as 60%).
Formula
saturation = Hill(current average adstocked spend) ∈ [0, 1]
Worked example
A channel at saturation 0.35 with marginal return 1.8× has room: spend can grow substantially before returns flatten. Another at 0.85 with marginal return 0.6× is spent through its curve — its strong average ROAS was earned on the steep early part, and the next euro lands on the flat part. The app marks any channel at 0.6 (60%) or above in amber as a scaling caution.
How Saldo Metrics computes it
The MMM passes each channel's weekly spend through an adstock transform and a Hill
saturation curve. The Hill shape is fixed at 1, which makes the curve concave —
returns diminish from the first euro, with no S-shaped slow start — and its
half-saturation point is set at the channel's median adstocked weekly spend; the
regression then fits how much margin the curve is worth. Saturation is that curve
evaluated at the
channel's current average adstocked spend, clamped to [0, 1], stored per run in
fact_mmm_channel_contribution and served from the latest run per method by
v_mmm_channel_contribution. It is the level on the curve where marginal return is
the slope: the two columns describe the same fitted point, which is why a high
saturation and a low marginal return arrive together.
Why it matters
Saturation is the why behind a weak marginal return, and the difference matters for action. A low-marginal-return channel at high saturation is a good channel fully exploited — hold budget, refresh audiences or creative to move the curve itself. The same marginal return at low saturation means the channel is weak per se, and the budget belongs elsewhere.
Common mistakes
- Reading saturation as a performance grade. High saturation often marks your best channels — you spent them up because they worked. It bounds growth, not quality.
- Treating 0.6 as a stop sign. The amber threshold is a caution to check marginal return before scaling, not an instruction to cut.
- Assuming the curve is fixed. Saturation is fitted against the audiences and creatives of the trailing window; new markets, formats or creative can shift the whole curve, and the next run will see it.
Where you see this in the app
Marketing → MMM, as the saturation column of the channel table, shown as a percentage — amber at 60% and above.