Inventory Optimisation means knowing exactly where to invest the next £1 of inventory in order to make the biggest difference to performance.
With profits increasingly under threat, the need for a highly optimised inventory is becoming ever more important at a time when a large number of organisations still use systems for managing their inventory that have one foot in the past. Even when sophisticated Enterprise & Resource Planning (ERP) or Sales & Operations Planning (S&OP) systems are in place, organisations can still struggle to deliver the anticipated gains.
Significant benefits to the business
Well optimised inventory can bring significant savings for the organisation by minimising working capital, reducing obsolescence costs (by only storing what is really needed) and simultaneously increasing revenue and customer satisfaction through improvements in stock availability.
Most business systems provide transactional services, but lack the sophisticated modelling and analysis that is used for optimisation. Even those with a degree of capability to provide stock optimisation may not have moved on much from their origins in the 1970s.
These systems at the lower end of S&OP maturity1 are employed for inventory management because of cost-savings, legacy issues or, quite simply, happen-stance as existing ERP systems have evolved to offer forms of inventory management. Whilst these systems lack capability, they are simple in operation and require little investment.
On the other hand, advanced inventory management and S&OP systems are armed with specific algorithms (or variations of) that have been developed to fit different scenarios and create optimum fit for the organisation’s unique needs and environment.
A myriad of options are offered to tailor the workings of the software, or the system may be offered with the promise that this optimisation will be performed automatically.
Having a large set of algorithms to-hand presents significant complexity to the organisations and for this reason can actually lead to sub-optimal performance. Even if (or particularly if) the algorithm is chosen on an automated basis, it will usually still require its own set of accompanying parameters to be defined in order to operate at an optimal level. Get these parameters wrong and the response to change can either be glacially slow, or it may appear to be much more random.
Different algorithms and parameters can produce significant differences in the forecast for particular Stock Keeping Units (SKUs). It is also not unknown for SKUs to move back and forth between algorithms, causing ripples of peaks and troughs through the forward schedule. When replenishment systems are being used to produce reliable schedules, this situation is far from ideal.
High configurability, low transparency
The range of parameters that need to be set, and the fact that the choice of algorithm is a black box to most users, can lead to an inability to challenge results and a feeling that ‘the system knows best’. With a well-optimised system this may be true, however many organisations lack the skilled resources or time to maintain the system parameters on an on-going basis. The parameters may be set once and forgotten, which is not ideal in a constantly changing world.
As a consequence of this, organisations may see high levels of inventory, but low availability - and they may not know how to address it. In extreme cases, since the results are clearly not right and no-one fully understands the system well enough to do anything different, this can lead to significant time wasted tweaking settings, making manual stock interventions, or creating off-system workarounds. A culture of fire-fighting can develop, where a significant amount of time across the organisation is spent waiting for or chasing stock. Is there scope for a different approach?