Procurement Data by the Numbers: Signs of Troubled Processes

This guest blog was contributed by Mahesh Kumar | 11/15/13
A large part of any enterprise budget is dedicated to acquiring and managing the technologies that run the business. But the processes for procuring that technology are deeply flawed, which my colleague Archana Penukonda discussed here recently in “Fixing a Trillion-Dollar Business Process.” How do we know the scope of the problem? By looking at the data that is the artifact of those purchasing processes – the purchase order.

The PO is the heart of the procurement process, yet the data from POs feed systems and processes that support ongoing planning, cost controls, audits and compliance. Errors and problems in the PO data can have a far-reaching business impact.

BDNA analyzed 7,500 technology purchase orders from four major enterprises in highly regulated industries throughout a two-year span. Because technology is critical in these industries, one might imagine that their purchasing processes would be more disciplined than other industries. And perhaps they are – but we found that a full 60 percent of the POs were incomplete or incorrect in some way.

What happens in the PO sometimes stays in the PO

Every purchase order tells a story. A deeper look at this set of purchase orders told a story of how the purchasing process was inherently flawed – often through no fault of the purchasing companies themselves.

Many of the problems fell into a few key categories.

— Missing quantities: In 10 percent of the purchase orders, quantities were embedded in the text-based description field – such as “Tivoli Asset Discover – 10 pack.” Quantity values embedded in text fields need human intervention to reach other systems that analyze and track entitlements. In this 10 percent of POs, critical quantity data is stranded in the purchase order.

— Maintenance complexities: Two percent of those POs included maintenance (a recurring cost) in the same line item as the license (a one-time cost). Because maintenance contracts must be tracked and managed separately, bundling the purchase with the licensing obscures important data.

— Inconsistent names: How many ways can you designate “Microsoft Windows Server 2012” on a purchase order? Analysis uncovered that most technology products had an average of nine different naming variations. 

— Inconsistent licensing: The same product is often sold in multiple licensing models – often to different parts of the same enterprise. In BDNA’s analysis, SAP’s Enterprise Server took the record for licensing complexity with 12 different licensing models.

How do you fix procurement processes?

It’s tempting to see this as a process problem; give people better training on how to fill out purchase orders, and your problems are solved. But the technology environment itself makes it nearly impossible to get the data correct.

Why? There are too many players. Some IT organizations attempt to standardize on certain core-set technology vendors. But even in tightly controlled IT shops, there’s always a “long tail” of technology vendors. For example, analysis of the server technology alone at a global manufacturing company found a shining example of the oft-cited 80/20 rule: 80 percent of servers came from four key vendors, while the remaining 20 percent came from an additional 27 vendors.

On top of this, there are no standards. Every vendor has different strategies for maintenance, licensing, naming and bundling. In addition, many organizations may acquire the same technologies purchased through different sources (manufacturers, resellers and VARs) – leading to additional layers of complexity and inconsistency in purchasing data including:

— Different SKUs or part numbers for the same equipment

— Different product naming/bundling

— Varying maintenance contracts for the same software

Change is the only constant. The technology sector is addicted to change, always chasing the next new thing. Vendors acquire each other, rebrand and rebundle solutions. Changes to product names or licensing terms lead to inconsistent product names in purchase orders that feed downstream business processes.

Data is the source of the technology procurement problem and, ultimately, the heart of its solution. By proactively cleaning and fixing the data in the procurement process, businesses can better understand and optimize their procurement processes while creating a strong foundation for ongoing operational and strategic planning, cost controls and compliance.

Mahesh Kumar

Mahesh Kumar is the chief marketing officer at BDNA Corp., a Mountain View, Calif.-based Data-as-a-Service solutions provider. He believes in big ideas that have ubiquitous application. A passion to democratize IT information led him to conceptualize, build and market the industry’s first Configuration Management System, the information hub that drives IT processes. At Kontiki, Kumar marketed products that provided anytime, anywhere access to rich digital content. He also made key contributions at Loudcloud, the cloud-computing pioneer. Kumar likes to golf, spend time with his family and venture on an occasional mountain climb. Kumar has an MBA from the Wharton School and a master’s in engineering from Clemson University.