This is joint post with Ananth Iyer, Susan Bukeley Butler Chair in Operations Management, Purdue Krannert School of Management.
USAID’s Global Health Supply Chain RFP is set to be one of the largest single awards ever made by the agency and final proposals are due in March 2014. Under the RFP, USAID very smartly provided data on some measures of the past performance of the USAID-funded supply chains that purchase and deliver life-saving health commodities like family planning methods and anti-retroviral medicines to developing country or other recipients. The data includes both the SCMS (PEPFAR) and DELIVER projects (see Amendment 3, fourth link to spreadsheet), under which about US$500 million of funding was spent annually between 2005-2013. This is the most data on the supply chain that has ever been disclosed in the public domain by USAID, and represents a resource for all.
However, the data are incomplete and these omissions may prevent better understanding and improvement of supply chain performance. Here’s the situation:
The USAID data includes a file describing purchase orders issued and associated shipments. The data provides detail regarding order flows from the suppliers to the warehouses in Singapore and the Netherlands, from the warehouses to customers and from suppliers to customers. It includes, for each order, the purchase order date, order available date, actual shipment delivery date, desired delivery date and an average lead time across all shipments by country. It also includes time stamped data regarding warehouse inventory.
But a crucial piece of data that is missing is the “shipment date” for orders. In the absence of the shipment date, it is not possible to identify the lead time between order shipment and delivery, a basic parameter required to understand the performance of a supply chain. Thus the logic for the quoted average lead times cannot be linked to the transaction level data. In the absence of outgoing shipment information, the product inventory at the warehouse does not reconcile with the inbound and outbound shipments, thus one cannot understand if orders placed are available for shipments (note that for orders to the warehouse, the order available date is missing).
It is tough to model the past and future functioning or efficiency of a supply chain if we don’t know how long it takes for commodities to get from point A to point B. It is also worrisome that the dataset lists an average lead time for some country pairs without any transactions, and leaves out such data for others for which there are transactions – leading the observer to wonder about the source of this data.
Finally, the product cost data is left out, thus preventing any analysis of the costs associated with the flows through the system.
This kind of data is standard for any assessment of supply chain performance, whether Walmart or USAID, and their omission reduces our understanding of the performance of flows managed on behalf of USAID.
USAID has also issued a separate call for proposals for business intelligence and analytics on the supply chain (search under www.ebuy.gsa.gov, accessible with registration). We hope to see better data and analysis as part of this new phase of investment in commodities and supplies, the largest and most significant use of US funding for global health.