Glossary
Counterfactual (promo ROI)
An estimate of what would have happened without a discount, built from matched cohorts (store, weekday, basket band, category) and reported with a 95% confidence interval and an 'observational, not causal' label. Used to decide whether to hold or add promo budget.
Why it matters for operations: a discount can look successful because sales rose, when they would have risen anyway. A matched comparison estimates the sales without the discount, so spending on promotions is judged on what it added.
Example: a weekday promotion is compared with matched cohorts of the same store, weekday, basket band and category. The result is shown with a 95% confidence interval and labelled observational, not causal, so the decision to hold or add budget is made with the uncertainty in view.
Where this comes up on budally.com
- Dutchie alternatives in 2026: switch POS, or keep Dutchie and add a layer
- Loyalty & wallet
- Discounts vs repeat rate: run the counterfactual before you add to Q4
- The loyalty gap: where a dispensary loses members between visit two and visit three, and the compliant way to win them back
- What the April 2026 Schedule III order changes for dual-license books.
Related terms
- Holdout
A randomly selected share of an audience (BudAlly default 10%) that does not receive a campaign, so lift can be measured against it. Campaigns ship with one; lift is reported with a confidence interval or 'not significant'.
- Rule leakage
Discount redemptions that fired outside the rule as configured (wrong weekday, wrong category). Computed from ticket attributes; the fix is drafted as a POS rule change for a person to apply.