Sales, of course, needs to see forecast accuracy at the level at which they give input. This is the only way they can improve on their inputs. The same is true for any other source of collaborative input.
The accuracy of the statistical forecast should also be measured at the level at which it is generated. The demand planner is responsible for examining these accuracy measurements and determining whether a given product has reached the level of maturity and stability at which a statistical forecast is adequate, minimizing the need for manual inputs.
Fighting for a Single Set of Numbers
It is also the job of the demand planner to fight for a single set of numbers. Since the demand planner is the gatekeeper of the forecast, she needs to lead the fight for that forecast to drive all downstream planning. Nothing is more meaningless than having a demand planner and multiple collaborators putting dozens of hours into generating a forecast if production planning or execution is going to ignore it anyway. It takes diplomacy as well as determination – and, most of all, producing a believable forecast – to ensure that multiple spreadsheets with their own forecasts don't sprout like weeds.
Knowing How, Why and What Demand Needs to be Monitored
A key skill for a demand planner is the recognition that his responsibility does not end with generating a monthly forecast. On the contrary, monitoring actual demand versus the forecast throughout the month is equally critical in order to better serve key customers as well as to recognize when a change of plan is needed. This is an area in which new analytical tools and dashboards can indeed provide great benefit – but it still takes a skilled demand planner to interpret what he sees and to act accordingly.
In addition to keeping an eye on demand as a whole, the demand planner must know which demand needs special attention as the month progresses:
- All forecasts are not created equal. Forecasts for key customers who regularly order at the end of the month, for example, need to be "protected" when supply is short. Customer segmentation and some sort of "forecast entitlement" check by customer service reps before an order is accepted need to be implemented and updated frequently to ensure that the best customers get the best customer service. In addition, high-volume, high-variability customers must be monitored closely, since their forecasts are, by definition, less accurate than others and a large unexpected order may upset production plans.
- A forecast is not just a number but also a range. A good demand planner recognizes that there is an implicit tolerance range around a forecast. She looks at the big picture and avoids knee-jerk reactions to deviations that are normal "noise" and require no change in plan. Conversely, she also needs to know when a deviation is large enough to be significant and to necessitate re-planning in conjunction with the production planner.
- How is the month going to end? Increasingly, management wants to know earlier and earlier in the month whether or not the forecast will be met. A skilled demand planner will have an "order progression" system to allow her to project early in the month, based on historical and current month order and shipment rates, what the final monthly demand will be. A forecast will never be perfect, but with proper analysis and the right enabling tools, a good demand planner can come closer than ever before to predicting at least a few weeks into the future.
Knowing When To Change Horses
A final skill essential for a demand planner is the ability to recognize when macroeconomic changes require an entirely new approach to forecasting. In severe downturns, businesses are generally slow to reduce their forecasts. At best, they may reduce next month's forecast but leave the outer months at unrealistically high levels. Inevitably, this results in inventories being far too high.