Optimization I: A Powerful Tool

Originally posted on on the e-Sourcing Forum [WayBackMachine] on Friday, 25 August 2006

Even before Aberdeen came out with its “Success Strategies in Advanced Sourcing and Negotiations: Optimizing Total Costs and Total Value for the Next Wave of e-Sourcing Savings” in June of last year, some of us already knew that decision optimization was the future of strategic sourcing. Moreover, the fact that they determined that the application of optimization tools to analyze total costs, and of flexible bidding functionality to uncover creative supplier solutions has enabled early adopters to identify an average incremental savings of 12% above those that basic, price-focused auctions alone have generated was no surprise to those of us who had been developing such technology, and monitoring its implementation success, for many years. That’s why innovative sourcing companies like Iasta (your e-Sourcing Forum blog sponsor) were already working on Bid Optimization capabilities (with version 1.0 released in December of last year) and focussed optimization companies like CombineNet have been pursuing improved optimization technologies and algorithms for over a decade.

And as you read this, I can tell you that Iasta is investing heavily in the research and development of Decision Optimization 2.0, which it expects to complete by the end of the year. Decision Optimization 2.0 will be based on the theory of Total Value Management (TVM) and continue to run on market leading solvers such as ILog’s CPlex. TVM models attempt to go beyond the capabilities of LCO (Landed Cost Optimization) and TCO (Total Cost Optimization) models by capturing the value, and not just the cost, of an award. They support qualitative constraints, to allow you to ensure the award will meet your physical constraints (durability, reliability, timeliness, low defect rate, etc.), and allow you to capture the impact costs associated with an award (such as marketing value, low return rates and high customer satisfaction, etc.) through constraints, fixed costs, and adjustments. TVM is the next logical progression in sourcing cost modeling (and an extension of the TCO modeling capabilities that were found in many previous generation modeling tools, which included MindFlow). But I digress.

Many innovative service and solution companies in the e-sourcing marketplace have been betting for the last five years (or so) that optimization is the wave of the future, but the vast majority have met with limited success (often surviving by M&A, like MindFlow, as pointed out by David in a post earlier this year) and many more are out of business.

Furthermore, the companies that have succeeded, have done so primarily due to acquisitions and other strengths. For example, Ariba acquired many of its customers from Free Markets and its customers praise them for their market knowledge and end to end platform capabilities that support integrated best practice processes from start to finish. Emptoris essentially doubled its customer base from the Dicarta merger, acquired many of its initial customers from its auction capabilities, and retained them through its own end to end platform, beefed up by many acquisitions over the years. i2 just isn’t a name I regularly hear in any sentence that contains “strategic sourcing” and “decision optimization”, and many companies that have survived, like SCA Technologies, are still relatively small in terms of customer base.

MindFlow, once acknowledged by the analyst and research groups, including AMR and Aberdeen Group, as the provider with the most comprehensive self-service platform for decision optimization, especially for CPG and Food Service, with its complex modeling capabilities and numerous constraint categories, is now virtually non-existent since the acquisition. In fact, when you get right down to it, the only company that has been around for the long haul and succeeded on optimization alone is CombineNet, and it has historically made most of its inroads in logistics and transportation, not strategic sourcing award allocation (although its customer base and focus is broadening). In fact, in a recent web search, the only recent news of significant note for 2006 that I could find in the strategic sourcing optimization arena, is the announcement in January that CombineNet is partnering in a joint venture with the University of Pittsburgh Medical Center to provide advanced sourcing solutions to the healthcare industry, as per this press release.

It is well known that many of the market leaders, and many of the big companies, are using decision optimization as (a critical) part of their strategic sourcing processes, but, considering that leaders typically make up less then 20% of the market (and that I know for a fact that not all market leaders are using decision optimization), it’s fair to ask “Who else is using decision optimization?”.

More importantly, when research has shown that repeated auctions on the same category quickly lead to diminishing returns, and often to net 0 returns after only 3 or 4 auctions, and that decision optimization often leads to incremental savings of 12% above and beyond other savings opportunities, why aren’t more companies, especially the mid-market enterprises, making regular, constructive use of this technology?

I believe the lack of use of optimization across the board is the result of numerous factors that have combined to downplay the appeal and importance of this technology over the last few years when concentrated efforts should have been made to introduce this technology as an overall component of any value-based strategic sourcing process. Tomorrow, I will discuss those factors and Sunday I will discuss why I think the time of decision optimization for sourcing analytics is finally here.


For a more in-depth discussion of decision optimization, what it is, what it is not, how it enables decision support, the benefits it provides, and strategies for success, see the “Strategic Sourcing Decision Optimization: The Inefficiency Eliminator” wiki-paper over on the e-Sourcing Wiki [WayBackMachine].