Global behavior parameters

weekly_traffic_multipliers

Positive daily scale factors that repeat across the week. They model the common observation that shopping attention changes by weekday. A multiplier above one raises expected traffic; one is neutral; a positive value below one lowers it.

daily_traffic_trend

A non-negative day-to-day growth factor. It represents gradual change across the report period, such as an expanding campaign audience. It should remain small enough that a short simulation does not become unrealistic.

conversion_probability_daily_noise_standard_deviation

The spread of daily random movement in conversion log odds. It models unobserved short-lived influences. Larger spread creates more volatile conversion probabilities without automatically raising their intended baseline.

performance_volume_noise_standard_deviation

The spread applied to daily touchpoint delivery volume. It prevents every day from having identical impressions and clicks.

path_audience_noise_standard_deviation

The spread applied to daily path audience size. It lets aggregated path counts vary separately from performance delivery counts.

performance_revenue_noise_standard_deviation

The spread of positive revenue variation for platform performance rows.

path_revenue_noise_standard_deviation

The spread of positive revenue variation for path outcomes. It is separate because the path report and platform diagnostics represent different reporting processes.

additional_unit_probability

The chance that a platform-reported purchase contains one extra unit. This distinguishes purchases from units sold.

repeat_purchase_probability

The chance that a converted path user contributes an additional purchase. This distinguishes converted users from purchase count.

Noise creates variation, not evidence. Private values should be calibrated from appropriate data and stress-tested across several seeds.