Improve the resilience of business plans using statistical confidence intervals
Generally, Sales & Operations Planning (S&OP) decisions are made based on a “best guess” or single point value forecast, rather than a range of possible outcomes (the statistical confidence interval). Even though planners and leaders understand that they have oversimplified the demand (and other) forecasts to create concrete plans, they might now know there is a better way.
Due to system limitations and process inefficiencies, you may see:
- Planners generating point forecasts to predict future demand (for example, we expect to sell 100 units in December). This is a deterministic approach and does not account for uncertainty or variability in demand.
- Planners producing range-based forecasts to better understand the confidence level around future demand (we have a 95% chance of selling 100-110 units and a 5% chance of selling below 100 or above 110), but S&OP leaders are still extracting a point value from the forecast to build a plan that can be used by the supply team.
Unfortunately, as soon as a point value is committed to the plan (eliminating the confidence interval), half of the value of the forecast is lost. In a world where planning activities can easily be unified in a cloud environment and automated with machine learning algorithms, the need for this oversimplification is trumped by the need for resilience to change.
A Practical Guide to Supply Chain Planning Maturity
Plans that take the confidence intervals into account can brace themselves for market changes when confidence is low (larger interval) and remain lean when confidence is high (smaller interval). This means planners can worry less about forecast accuracy (as we know – all forecasts are wrong) and focus on the level of confidence around the forecast (statistically speaking) and make the best decision.
“In business, range-based S&OP stacks the odds in your favor, and ensures no-surprises and no-excuses performance regardless of demand outcome,” according to Blake Johnson, Consulting Professor, Stanford University.
Understanding and incorporating the confidence intervals in S&OP, inventory and supply plans is a key component of protecting the supply chain (and its customers) against demand variability, potential stock outs, and product obsolescence.