Extended Supply Network Information Models: The Missing DNA in the Digital Supply Chain (Part 1 — A Backstory)

supply networks Andrey Popov/Adobe Stock

Let the reader be forewarned, this story will meander. Now that you’ve been warned, let’s begin.

When I was an engineering student in college, back in the late ‘80s, I decided that my senior thesis project in operations research should be to redesign the university mail system delivery route on campus. This required me to walk around campus for endless hours measuring the distance between streets and delivery locations (sort of a very old school equivalent to the Google GPS mapping cars) to build a supply network optimization model that I used to re-jigger the route and save both time and distance for the campus mailman. OK, it wasn’t exactly earth-shattering, but it “set the hook” for me about how software-based models could be used to solve real world operations problems.

A few years later, in the early ‘90s, I entered the wild and wacky world of management consulting and began to play around with other sorts of cool systems such as factory simulation, engineering data management (which morphed into product lifecycle management), MRP II (MRP with a feedback loop!), object-oriented back office workflow simulation and, most fun of all, packaged supply network design and optimization. I performed a supply network optimization project for a garden equipment manufacturer and came up with various optimal network configurations based on some scenario planning. The lights would dim when my poor desktop server had to run these optimization routines (I had a lot of 45-minute coffee breaks waiting for the solver to eventually pop out the answer) but the results from the analysis saved 16%, and the best part was the ability to visually show how the supply network flows could be redirected. Still, this class of tools has been fairly limited in terms of only focusing on the internal supply chain, such as warehouses (and inventory positioning within the network) and some manufacturing sites.

Eventually, the fates of consulting led me to some client work in strategic sourcing, which was red hot in the mid-to-late ‘90s. I liked the analysis and modeling of spend, costs, markets and ultimately, supply strategies. But there was something missing. The strategic sourcing work was, in essence, an episodic workstream to simplify the extended supply network by rationalizing suppliers. But, it didn’t really model the supply network deeply and look at the supply network from a deep TCO and value-beyond-cost standpoint — it was still biased heavily toward purchase price savings.  

When the world got a little nutty in the late ‘90s and early 2000s with reverse auction software from providers like Tradex and FreeMarkets (now SAP Ariba) sprouting up, I joined AMR Research in 1999 (now Gartner). As a side note, one of the analyst relations people from FreeMarkets was a young fellow named Jason Busch, and he wasn’t happy about what I was writing about the firm’s nascent efforts at software development — but that’s a story for another day. Also, at that time, there was a folksie blogger from New Hampshire named Debbie Wilson, who ran a “Cool Tools for Procurement” blog. She in turn filled Mickey North Rizza’s shoes when Mickey filled my old role when I left to join The Hackett Group, in 2004.

Anyway, back in 1999, there were no real “industry analysts” in procurement, only in areas like ERP, supply chain, CRM, PLM and so on. Since my background was supply chain, my context for procurement was for it to be part of a broader value chain rather than just an n-step sourcing factory for churning out competitive bidding events. Actually, I was just recently looking back at an e-sourcing market evaluation study I led back in 2002, and interestingly, most of the tools currently on the market still don’t score well against those requirements — which I’ll return to in a later post.

The one area that did hold promise from a supply chain standpoint was the area of bid optimization. Pioneered in the 1990s, it was limited to freight bidding but began to creep out into direct materials, packaging, MRO and other areas as providers like Emptoris (IBM), MindFlow Technologies (IBM) and CombineNet (SciQuest) began to apply expressive bidding techniques (or “market informed sourcing”). Today, firms like Trade Extensions and Keelvar are continuing to push the envelope here. The reason why I have been passionate about this area (although my colleague Michael Lamoureux is even more so!), is not just because I view this type of sourcing strategy as akin to corporate strategy, but also because it sort of embodies a design ideal of optimized real-time extended network design.  

That's a mouthful isn't it? But, think about it. It’s basically a process to dynamically reconfigure suppliers, internal operations (i.e., a factory can be modeled as a supplier in a make versus buy analysis), freight assets, packaging, supply chain financing and so on. If you want to dive into this area, you can access some level of insights from the deepest research I’ve probably done (back in 2003) in the area of “strategic transportation sourcing” (contact me if you want to chat on the topics). In fact, I know of a global freight firm that uses such a bid optimization tool to model its complete network based on expected shipping demand.

At this point, I hear you saying: “OK, Pierre, this is a nice story, but what’s the bottom line?”

The bottom line here is that a supply network data model must live at the heart of any supply chain, whether that model is used for projects in supply network design, bid optimization, supply risk management, supply market intelligence or other areas. The problem is that no single solution or even class of solution offers up this robust data model that can be used in so many areas.

But this is starting to change. In the next part of this series, I will outline how a multitier supply network data model is critical to all supply-side processes and application areas, and how the required data model implicit in running a true global supply network is as fundamental as the data model change required to move from an on-premise traditional enterprise class application to a true software-as-a-service (SaaS) multitenant application. Stay tuned!

Share on Procurious

Discuss this:

Your email address will not be published. Required fields are marked *

This site uses Akismet to reduce spam. Learn how your comment data is processed.