A familiar situation

What is your response to risk in business? Would you buy an IT system with a high risk of failure? An article by the consulting firm McKinsey considered that more than 60 percent of supply chain planning IT transformation implementations do not achieve anticipated business outcomes and take more time and/or money than expected.

This figure may be surprising, but is similar to the situation in past eras with implementations of ERP and MRP systems in businesses. Why has the situation not changed? It starts with thinking by technology providers and internal IT departments that because the various applications run on computers, then an application with a three letter acronym is an IT project, not a business improvement project. This thinking is enhanced by the desire of senior management for making departments (vertical silos) more efficient, directed by centralised control of the business, which requires ‘integrated’ systems that provide continuous updates of data and information.

However, the supply chains are a business process that flows from the customer’s customers to the supplier’s suppliers, so the objective is for supply chains to be more effective. Supply Chains are not a function, within a function based organization structure, but a Flow. The core functions within the Supply Chains group are Procurement, Operations Planning and Logistics.

Opportunity for failure

The availability of Cloud computing and Artificial Intelligence (AI), means that the upfront cost for what is called ‘IT transformation’ of supply chain planning is reducing. The ‘transformation’ uses Enterprise Resource Planning (ERP) tightly linked with Advanced Planning Systems (APS). Combined with the misunderstanding about the role of supply chains, the software provides potential opportunities for failure in the implementation.

A continuing challenge in the world of supply chains is definitions and even meanings. For example, APS has been referred to as: Advanced Planning Systems; Advanced Planning Solutions and Applied Planning and Scheduling and there maybe others. APS contains software tools that enhance planning systems. These include mathematical algorithms for demand forecasts; to optimise product mix and sourcing plans, and to plan and schedule production and distribution with constraints. However, APS tools are built using linear optimisation technologies to address complex, non-linear supply chains.

As companies become larger and maybe multinational, their data sets become increasingly complex, with dependencies that are difficult to comprehend with lots of transaction level data. These issues can demand a ‘better’ planning system, with a higher level of decision support in Operations Planning, and more planners and schedulers. Or do they?

Integrated vs Connected IT Applications

By definition, ‘integrated’ supply chains are linear in concept and data flow is supposedly continuous, with minimal latency. The core data is within applications designed for vertical processes, such as: Enterprise Resource Planning (ERP), Customer Relationship Management (CRM), Supplier Relationship Management (SRM), Warehouse Management System (WMS), Transport Management System (TMS) etc.

These applications are designed to improve the efficiency of their functional processes, but a focus on functional optimization can upset the balance of supply chains, with a potential increase in inventory and total costs. But even when tightly integrated, the resulting system is not designed to improve the effectiveness of supply chains. And there is no incentive for software providers to enable interoperability across the networks of trading parties.

As noted in the previous blogpost, there are many issues with master data that require a near continuous audit processes. Also, the MRP logic within an ERP system requires routine (and expensive) maintenance to keep the applications operating efficiently. And the ERP system also requires a structured process for routine updates.

In comparison, the concept of Connected supply chains has a focus on Horizontal flow processes, such as: Revenue Management, Sales and Operations Planning (S&OP), Inventory Management, New Product development and launch and Supplier development. These flows require a capability for the streaming and pooling of data that enables the schema to be created only when reading data that has many unknowns from a stream of new sources.

An Integrated or Connected approach must recognise that Supply Chains Planning requires a good understanding of supply chain principles. Also, the models used in a system must be industry, or even company specific. This requires deeper understanding, knowledge and application insights by the people assigned to the implementation project by the manufacturer, software supplier and consultants.

And this leads to the underlying challenge of providing complex solutions for complex problems. The more complex the challenge, the deeper is the reliance on a shrinking pool of knowledgeable and capable people, who unfortunately bring with them accumulated prejudice and bias. Instead, an alternative approach is simplification.

A different approach

Large, centralised systems for supply chains planning are more likely in global companies that perceive a need for a corporate level visibility of operations. However, planning regional supply chains (within a geography or country) will have less complexity if production is mainly destined for the region. Because it is likely that a high percentage of production and logistics activities have been outsourced, Supply Chains Planning is more likely to be a region business network planning process with S&OP conducted at each factory or country DC for imported products.

The development of APS attached to ERP systems has focussed the planning processes on being precise; with the resulting plans being precisely wrong. Instead, the focus must be to understand the probability of demand and other events. Consideration of data in S&OP does not commence until after the ‘freeze’ period and continues out to the time required for enhancement to capacity, so the probability of a data range between optimistic to pessimistic is the best option. This data can be used in ‘what-if’ modelling within S&OP to provide scenarios of future events.

As important is understanding planned time, variance and latency within each defined supply chain and how they apply to meeting a customer’s needs. Each customer order has latency – the time to translate customer sales to market sales then into re-order points through the sales channel back to the manufacturer. There is a need to understand constraints in supply chains and lead time variation and requirements for inventory buffers. And to measure in-transit inventories (especially if between regions).

So, a different approach is to plan with customers as the driver. Focus on simplification of processes to reduce the cash-to-cash cycle and design a customer response for each supply chain. Reduce organisation complexity and emphasise the responsibility of individuals for decisions. And do not subscribe to the approach of implementing complex solutions for complex problems.