1. Background
Although APO is not part of the implementation in this phase and demand planning will be done outside SAP system, demand planning process has been defined as part of Phoenix due to its being the key input to business decision process. It is the start of business operation and its quality will highly impact other processes in the chain.
2. Objectives
To review Osotspa As-Is demand planning process versus industrial best practice, identify and recommend areas of improvement.
Currently from back order root causes analysis report Jan to Aug 2016, actual sale higher than forecast is the number one reason of back order with around 40% contribution followed with supply planning and replenishment as the second reason with around 20% contribution.
With demand planning process improvement, it should be aimed for back order reduction due to under forecast by xx%
3. Findings
3.1 Organization:
In personal care business unit, role of demand planning is clearly defined and separated from supply planning according to best practice. However, major task that has occupied time of demand planners is to prepare sale bottom up templates drilled down to detailed route level. In total it is around 60 templates to prepare in each month. This causes demand planners having not enough time to do demand driver analysis.
Regarding resource and workload, currently there are 3 demand planners plus one assistant demand planning division manager against 900 forecasting SKUs. Business is under SKU rationalization process. Demand planners’ workload will depend on number of SKUs and customer forecasting level, so can be decided enough or not enough when SKU rationalization is completed and forecasting level responsible by demand planners is agreed.
In beverage there is no demand planning role in the business unit.
3.2 Granularities of Forecasting (for personal care)
- Product: SKU
- Customer: By channels (MT, K, OMC) in Forecast Pro with MT drilled down to routes, K and OMC split into routes on Excel template using proportional factors from historical sale data 6 months back and updated monthly to split suggested numbers from Forecast Pro
- Time bucket: Month in Forecast Pro, month & week on bottom up template (refer to table 1 below)
- UOM: Carton (carton and converted total into Baht on excel bottom up templates)
- Level of review: MT – all SKUs month 1-6 with month 2-3 in weeks
K – Priority SKUs, month 1-2 in weeks
CJ – all SKUs month 1-6 with month 1-2 in weeks
OMC – priority SKUs month 1 in weeks, month 2 in month
- Forecasting horizon: 6 months