The phrase ‘closed loop’ has a variety of usages in supply chain management. Physically it is used in the context of returnable/reuseable transit packaging, or more generally in any context where reverse flows are significant (thus, manufacturers’ plans to cope with the End of Life Vehicles Directive might be described as ‘closing the loop’).
But closed loop is also used in a deeper sense by supply chain designers and system vendors to describe the integration of those two implacable foes, planning and execution. (or if you prefer, demand and supply).
Notoriously, the supply chain planning and execution functions in the typical organisation have had at best a nodding acquaintance – different teams, usually in different offices, sometimes on different continents. This wouldn’t in itself matter if the two functions’ systems and information flows were less like ships that pass in the night. As it is, the plan and its execution generally touch hands only periodically or after a particular plan (a promotion, for example) has ended and the chance for corrective action has been missed. Lessons may be learned but, like generals, we are always fighting the last war.
This problem has not gone unnoticed but until quite recently there was little that any but the smallest organisations could do about it. Real-time information on execution performance was scarce or non-existent – problems only became apparent when and if a customer complained. Plans themselves could only look at supply execution capabilities and likely performance levels at the most macro of levels. Attempts were made to ‘integrate the enterprise’ but typically on the back of ERP or similar. As Sanjiv Sidhu, president of software vendor i2 pointed out at the recent i2 Planet congress in San Diego: ‘50 per cent of SCM data doesn’t exist on an ERP system. ERP systems weren’t designed to execute plans, they were designed to execute transactions.’
But another approach is now possible. IBM calls it ‘on-demand SCM’ while others talk of ‘demand-driven supply networks’ or ‘agile supply chains’. i2 calls the concept ‘closed loop’ – less sexy but more descriptive. Whatever, the point is that technology, in particular the shift toward web-service (a shift, suggests Sidhu, as significant as the previous generational move from central batch-processing on mainframes to the distributed client/server architecture) makes it at least theoretically possible to bring together information on supply chain performance in as near-real-time as makes no odds, in a way that allows corrective action to be taken.
But Sidhu’s more important point is that the technology by itself won’t achieve anything – it is all about how you structure yourself and your supply chain partners to use this breadth and depth of real information. His gospel – and he does get evangelical about this – is ‘Plan-Do-Check-Act’ (PDCA).
Sound familiar? It should. It is of course the basis of Total Quality Management which attacked with stunning success the issues around variability in product quality, and millions of people know all about it. You plan what you are going to do; you do it; you check what you have done; and then you act to correct any actual or incipient problem or trend. The approach has enabled even job-shops to achieve close to six-sigma levels of repeatability – that’s in the one defect per million range. By contrast, a supply chain that offers 99.8 per cent service levels, which most of us would regard as pretty darned good, has a ‘defect’ level of one in 500.
As Sidhu says: ‘If the enemy is now variability in operations, we need a similar process.’
We have the technology. We have bar codes and soon RDIF; we have GPS and other asset-tracking devices; we have EPOS and other genuine sales or demand data. We have relatively cheap devices and networks that can bring this information together, regardless of format so that it can be used. We have, increasingly and necessarily, methods of performing management by exception so we don’t worry about things going right. What is missing in many supply chains is the mindset that says ‘we can check our performance against the plan, – not when the plan is over to understand why we have lost money or customers, but every day to correct the execution or the plan before we lose that cash and those consumers’.
Like generals, supply chain managers are often enveloped by the fog of war. If a whole campaign like a sales promotion is a failure, that fact is often all we know. Every plan should have worked – the first day of the Somme should have worked yet 90 years later historians still argue over whether the plan itself was misconceived or whether a handful of individual ‘failures to deliver’ condemned the whole enterprise. Great War generals notoriously had almost no real-time information – but we do and we can do better.
Sanjiv Sidhu and many like him believe this is possible, indeed imperative. And Sidhu would go further – not just the plan-execution loop needs closing through PDCA methods, he argues, but the same principles must apply to areas such as supply chain design and strategic sourcing. Supply networks and sourcing relationships cannot remain static in a fast-moving world – nobody can afford to wait until the end of the contract to find out what has gone wrong. And if you do you will never discover why.
Briefing points
- The problems of product variability have largely been cracked. The same approaches should now be brought to bear on variability in process and operations.
- Affordable technology now exists to close the loop between the planning and the execution of supply chain management using real-time information.
- This requires supply chain stakeholders to implement a PDCA feedback process continuously while there is still time to change, rather than periodically or at the plan’s end.
- The same continuous feedback approach may also usefully be deployed in supply chain design and in sourcing activities.
