Baseline (sales without a promotion)
How much you would have sold without the promotion: from history, seasonality and trend.
Formula
Baseline = average (sold ÷ seasonal index) before the promotion × the week’s index
How to read
Without a plausible baseline there is no effect, only revenue. Check the estimate with a control group.
Where it is discussed
Topics
- Season, price and discounts
- Wholesale terms and credit
- Markdown and exit price
- Stock accuracy, shrink and availability
- Marketing budget and testing impact
- Discount mechanics and the promotion calendar
- Promotions: planning and evaluation
- Season results: brand, category, SKU
Calculations
- Promotion result: incremental profit
- Co-funding a promotion with a platform or partner: the brand’s share
- Checking an AI forecast against a naive baseline: WAPE, bias, improvement, share of periods and the gain net of costs
- A/B test on proportions: sample size, duration and reading the result
- Post-promotion dip and pull-forward demand
- Control group: promotion effect (difference-in-differences)
- Promotion post-mortem: plan, actual and verdict
- Promo effect without a control group: a baseline from last year
Documents
- Promo calendar and promotion effectiveness
- Descriptive statistics for retail: center, spread, outliers, intervals
- Demand forecasting: smoothing, seasonal indices, analogs and error control
- Price elasticity of demand: how to estimate it from data without being fooled
- A/B tests for promotions and promos: hypothesis, sample, power, p-value, peeking
- Holt-Winters, seasonal decomposition and Croston: forecasting with trend, season and intermittent demand
- Multiple regression of sales: demand drivers, advertising and promo contribution, multicollinearity, lags and model checks
- Bayesian estimates of conversion and buyout: small samples, prior distributions, shrinkage and comparison with the frequentist approach