As supply chain consultants, we often help clients navigate complex change–from adjusting to cost or competitive pressures to unique challenges like an M&A. These challenges all demand careful analysis of multiple financial and operational outcomes. There’s often a lot hanging in the balance, so it’s rarely a good time to “map out” supply chain changes or cost reduction goals on the back of a napkin.
Our most-successful analyses and recommendations are always built on models that use high-quality data–reasonably complete records that contain the breadth of data applicable to the task at hand and are validated for the client project’s purpose. The type of client data we employ in our analysis depends on the problem to be solved. But more on that in a minute.
“Garbage in, garbage out”: Not the right model
Data is as much a key ingredient of the supply chain as trucks and warehouses. And the more accurate information available about products and how they flow through a supply chain, the more accurate the model. But as Aberdeen Group research has determined, only 16 percent of companies have data quality that could be considered adequate, leaving 84 percent with subpar data quality.
That leaves an immediate problem for long-term supply chain modeling and planning:
- Typical supply chain challenges (like supply chain planning and strategic network design) often require very complex, data-dependent, periodical evaluation or even involve multi-tiered capital investment decisions, from building new or extending existing facilities to closing, reconfiguring or re-siting plants, distribution centers (DCs) or warehouses.
- Chainalytics’ fact-based approach employs software that depends on accurate data feeds to build complex models of a supply chain and its behavior over time.
It’s an industry truism that over 60 percent of a supply chain design process/project time is typically spent identifying, collecting and validating the data that powers the model and helps to create a view of the supply chain. A further 20 percent is spent analysing the model output.
How different kinds of data affect your supply chain results
All data is not created equal, when it comes to its applicability/usefulness in supply chain modeling and design. For example, master product, stock-on-hand, warehouse and demand data is crucial for designing warehouses, while master, production, volume, financial, qualitative and demand data is required to design supply chain networks:
- Master data includes data related to products and their relationships, for example, product groupings, units of measure, physical attributes, (weight and volume), units per carton, cartons per pallet, layers per pallet, etc.
- Inventory data includes inventory volume and value and allocations between DC’s in the network.
- Warehouse Data includes storage capacity, storage characteristics, logistics equipment, personnel and contracts (outsourced).
- Production Data includes product portfolio and production capacity.
- Volume Data includes inbound (suppliers to plant), plant to DC, inter-facility (stock transfer), direct deliveries (plant to customer) and DC-to-customer deliveries.
- Financial Data includes transportation, warehouse, production and inventory data.
- Qualitative Data includes customer service, lead times and business requirements data.
- Demand data, includes actual historical and forecast demand for months or even years and is critical in determining future growth requirements.
Each of these data sets powers some aspect of the physical supply chain, the financial supply chain (because we all like to get paid for the products we make) and the information supply chain. Of these three, information supply chains–and the consistency of data that is shared with and amongst partners– will continue to grow in importance, especially given growing collaboration amongst trading partners and increased government and regulatory product-related data requirements.
But if you’re facing supply chain change or a network redesign, it’s still important to keep the larger picture in view: Supply chains won’t succeed simply because they have the “best” information; they’ll succeed because they successfully link information management disciplines to specific and measurable business outcomes.
To get your process rolling, it might pay to do at least a cursory survey of how data is selected, used and managed in your company. After all, anyone who touches corporate data shares the responsibility for data quality, even if it is only to tell someone who can do something about it.
Richard Koch is associate director for Chainalytics’ Logistics competency, where he supports the firm’s continued growth in the Asia-Pacific region. Richard has over 25 years of supply chain consulting experience, including facility design and engineering, outsourcing and third-party logistics and supply chain network optimization. He is also highly experienced with software evaluation and selection as well as operations management, with particular emphasis on warehousing and transport optimization.