DIAG-6 · Diagnosis · Invariant — holds by construction, no expiry
Spend is endogenous to demand
The law
When budget is chosen using any signal of demand and spend affects outcomes, outcome-spend covariation cannot separate demand from budget policy; separation requires variation independent of the budget rule (exogenous demand signals, natural experiments).
Scope
Any account where budgets follow seasons, forecasts, or automated bidding, which is nearly all of them. Genuinely constant budgets are the exception that proves it.
What this licenses you to do
Correlate every fitted demand or seasonality curve against the brand's own spend series; reject any demand claim that cannot name an identification source independent of the budget decision.
Evidence
Observed building a production demand model: seasonal covariance looked like a valid demand-elasticity signal until tested against held-out data, which is why the shipped rule only trusts budget change-events and day-of-week jitter for this estimate, never seasonal covariance.