The Role of Forecasting in Demand-Driven Supply Chains

Back in the relatively stable business environment of the 1950s, it was difficult enough to achieve good forecast accuracy – how much more difficult is it in today’s VUCA (Volatile, Uncertain, Complex, Ambiguous) world? The answer, of course, is extremely difficult.

Today, 80 per cent forecast mix accuracy is considered ‘world class’, despite its hiding the fact that most SKUs in such a portfolio will be achieving less than 60 per cent (i.e. greater than 40 per cent error!) because their demand volumes are medium to low and their statistical variability is medium to high.

In the 1950s, when competition wasn’t as fierce as it is today and generous inventories for achieving reasonable service levels weren’t an issue, none of this was too much of a problem. But today there is unrelenting pressure to free up cash, achieve great service and reduce costs – all at the same time. Unfortunately, the means by which most companies currently manage their supply chains hasn’t moved on much from the high forecast accuracy dependent MPS/MRP logic that was ‘conceived in the 1950s, coded in the 1960s and commercialized in the 1970s’ (2). Even so-called ‘advanced planning systems’ and ‘inventory optimisers’ are still using this basic, but flawed, ‘forecast push’ process.

The inevitable result of using inaccurate forecasts to drive MPS/MRP is unbalanced inventories and service threats that cause, through the consequent avalanche of exception messages: backorder preventing expedites; schedule interruptions and fire-fighting that lead to use of unplanned capacity, excessive and variable lead-times, and excessive, but still unbalanced, inventories.

Fortunately, there is an alternative: Demand Driven Supply Chain Management (SCM) which is simply ‘enterprise(s)-wide pull’ that is now feasible where traditional visual-management-based Lean Pull isn’t: across complex and dispersed supply networks through use of a Demand-Driven software system that operates through ERP.

The essence of Demand Driven SCM is the deliberate strategic positioning, sizing and maintaining of independent de-coupled stock buffers across the supply chain and their ‘buy, make or ship to replace’ replenishment in line with demand – not the forecast.

Which isn’t to suggest that forecasting is no longer necessary!

Forecasts are still required for S&OP/capacity planning, exceptional and extreme advanced build event management and the stock buffer sizing. However, for all of these, there is no requirement for a high degree of item-level, time-phased forecast accuracy, so the forecast mix accuracy KPI can be dispensed with.

When Demand-Driven SCM replaces forecast push MPS/MRP, the benefits, due to minimization of supply chain variability are, transformational:

1. Achievement of planned service levels, from
2. Up to 50% of the average inventory, using
3. Significantly less capacity / higher OEE, with
4. c85% reductions in planning lead-times, all
5. Without expediting or fire-fighting and any focus upon accurate forecasting

Can your business afford to stay stuck in the 1950s?