How Expediting Destroys Factory Performance (& how to avoid it)
Nearly all ex-stock supply chain planners spend a great deal of time expediting and re-planning to head off imminent backorders. They do so because their time-phased (eg monthly or weekly bucket) sales volume forecasts versus stock position shows threatened use of safety stock or a deficit - often through MRP exception messages.
That this occurs so frequently is a symptom of the great weakness inherent in ERP/MRP/APS - reliance upon high levels of item forecast accuracy. Unfortunately, even when world class levels of forecast mix accuracy across a portfolio are being achieved (ie. 80%) the accuracy on the majority of items (usually those items with low to medium volumes that have proportionately higher levels of month to month variability) will be found to be less than 60%.
Despite MRP time-fences, the expediting, schedule interruptions and re-planning caused by this use of inaccurate sales forecasts and similarly inaccurate master production schedules, use high cost capacity and increase the lead-times and inventory of the interrupted items - both due to to the interruptions and the upstream material congestion that the interruptions cause.
Increased lead-times obviously puts the affected items at risk of service issues as they arrive late which sets off yet more expediting and a "service threat --> expediting --> longer lead-time --> etc" vicious circle results. Eventually a decision to permanently increase lead-time parameters and / or safety stock follows but the root cause of the problem (inaccurate forecasts and master production schedules) has still not been resolved so these companies, and their factories, end up operating with significantly excessive capacity, lead-time and inventory (the latter because of over forecasts, the safety stock increase and the increased lea-time parameter, see Little's Law (1) but still suffer backorders and periodically require over-time to catch-up.
Supply Chain Variability and Lean
Schedule interruptions are one of the 3 key sources of supply chain variability (which is effectively anything that disrupts the ideal of perfect material flow: ie. the smooth movement of material pieces through the supply chain in line with demand that would deliver perfect service/maximised revenue with zero static stock), the other 2 sources are batching and material movement disruptions such as processing machine break-downs and quality issues. In this light, Lean activities can be seen as those which minimise supply chain variability (eg. SMED / smaller batches, TPM, TQM etc - 2) and enterprise-wide Pull is the activity which contributes to the elimination of schedule interruptions, expediting and re-planning.
Pull
In contrast to the use of inaccurate forecast-driven master production schedules (ie. Push) as used by ERP/MRP/MPS, Pull uses re-order point / cycle mechanisms at selected positions through the supply chain that replenish what has been consumed, thereby helping to deliver Flow in that the end-to-end material replenishment is driven by demand and, therefore, is always accurate (NB. the item level forecast is still used: the re-order point / cycle inventory target is calculated using its average to account for trend, and it is also used for managing exceptional demand and supply events that require advanced stock builds eg. significant promotions, capacity constrained seasonality and factory shutdowns).
Many ex-stock manufacturing factories operate with localised pull mechanisms within the factory (eg. kan-ban, CONWIP, POLCA) but very few use Pull end to end - from distribution to RMs. That's because ERP systems are designed with a dependent demand network and cannot operate with more than one independent inventory position, nor do they have the essential inventory target sizing and concurrent capacity planning functionality (using the forecast) alongside the demand-driven replenishment/execution network.
Results
But companies that have adopted the demand-driven / e2e pull approach to replenishment across their sourcing / manufacturing / distribution networks, using the appropriate software (it often operates under the umbrella name Demand Driven MRP), benefit enormously from the resulting accurate flow of their materials and minimisation of performance destroying expediting and re-planning - typical benefits are:
Achievement of planned service levels, and
c40% inventory reduction, and
Shorter lead-times, and
Higher capacity utilisation / OEE (because of less schedule interruptions and more factory floor stability) and
More value-add from Planners with significantly less fire-fighting and stress
For more detail on why and how the demand-driven/pull methodology works (and forecast-push doesn't, no matter how much you spend on digitalisation / AI / real-time technology) see The Supply Chain Replenishment Problem (& how to solve it)
1 - Little's Law: inventory = lead-time (days) x daily throughput
2 - Hopp and Spearman, authors of 'Factory Physics' describe Lean as being "fundamentally about minimising the costs of buffering variability", those buffers being time, inventory and capacity.