Well, the doctor has news for you. You haven’t. In fact, you’re not even close.
You might be applying at least baseline optimization to the majority of your high-dollar and/or strategic categories. You might be in the Hackett Group Top 8%. You might be building Billion Dollar sourcing models. You might be years ahead of your peers. But the reality is that when it comes to true strategic sourcing decision optimization (SSDO) mastery, you’re not even close.
With the exception of the two e-CHAOS vendors, the doctor interacts and/or works with all of the remaining vendors who offer true strategic sourcing decision optimization (which isn’t a hard thing to do as there are only seven*1  vendors in total with a solution that meets the minimum requirements as set forth in the wikipaper), knows the depth of the projects these vendors have supported, and can say with confidence that the best of the best have barely mastered the basics of optimization 2.0. Barely. And optimization 3.0 is on the way.
[ As a history lesson, optimization 1.0 was circa 2000 when the first solutions that minimally met the four basic requirements of solid mathematical foundations (MILP), true cost modelling, constraint analysis, and what if? capability hit the market. Most of these were basic, supporting only supplier – product – customer DC mappings; unit and transportation costs and then one level of discounts or rebates; capacity, allocation, and min/max supplier selection constraints for very basic risk mitigation; and manually created what-if scenarios. In addition, maximum model size was limited, large models took hours to days to solve, and setting up and importing all of the data from multiple bid sheets across multiple spreadsheets often took days.
Then, circa 2005 to 2007, as a result of a considerable increase in computing power, algorithmic improvements, and domain knowledge, a few solutions started to improve rapidly and we hit the beginnings of optimization 2.0. The platforms evolved to make full use of the theory of logical variables in the MILP solvers; they also supported multiple supplier locations, product substitutions, and differential costs by lane*2; a buyer could define costs by way of a cost model with as few or as many factors as desired, at multiple tiers and with volume or spend-based discounts; a full plethora of allocation, capacity, and risk mitigation constraints that could define required and desired splits, address risk mitigation or mandate awards to a set of products, suppliers, and or regions, etc.; and could automatically generate what-if scenarios based on automatically adding or dropping previously defined or newly defined constraints, historical versus current pricing models, and other factors. In addition, import and export was streamlined from RFX, Auction, spreadsheet templates, and ERP systems (where standard transportation and overhead pricing was kept). State of the art report generators and OLAP capability was integrated so that not only could you generate scenario reports and comparative reports across scenarios, but you could also dive in to see what was driving the savings against the current sourcing strategy and, more importantly, what was driving the costs compared to the unconstrained baseline scenario (and zero-in on what business rules might be too costly). ]
The reality is that the average best-in-class organization is only doing T-CAP strategic sourcing decision optimization, and is still far from achieving TCO. Basically, when the average organizations build their cost models, they are focussed on the costs of acquisition and production (and distribution) of the goods they are buying. They’re not incorporating downstream maintenance, service and return costs and not considering end-of-life reclamation, recycling, and disposal. Nor are they breaking the acquisition cost models down to determine the upstream impact costs associated with the supplier or production method. For example, if the supplier runs their factory on dirty coal and the company has pledged carbon neutrality and has to buy carbon credits to achieve their goal or the working conditions in the factory are unhealthy (and the factory would be closed down if it was in America) and this adds more fuel to the fire of the CSR activists and is costing your organization brand value, these costs also need to be considered. As a result, the organization is capping its potential return from optimization. Not only is the organization not achieving TCO, but it’s no where close to achieving TVM (total value management), which is what it has to achieve if it wants to realize true optimization 2.0 mastery and move on to optimization 3.0.
And the average organization is not even thinking about the more advanced opportunities that the next generation 3.0 capabilities will enable. Right now, the leading strategic sourcing decision optimization vendors are integrating new capabilities in the new versions of their products that are currently in development, with some basic 3.0 capabilities already released! The convergence of big data, advanced analytics, and decision optimization into a single platform is enabling a host of new capabilities that the average organization has not yet envisioned, including the 6 next-generation advanced sourcing optimization capabilities outlined in Sourcing Innovation’s new white paper on Optimization, What Comes Next (registration required), sponsored by Trade Extensions (which is one of the vendors working hard to give you tomorrow’s optimization solution today).
Companies that master the 6 next-generation advanced sourcing optimization capabilities described in Optimization, What Comes Next (registration required), will not only be the first to master optimization 2.0, but will be the first to enter the world of optimization 3.0 and find savings and cost avoidance opportunities that they never even knew existed.
Are you ready to crank the amp and take it to 11? If so, download Optimization, What Comes Next (registration required) today!
*1 search the SI archives if you don’t know who the seven are
*2 the MindFlow Model, which was recognized as the first SSDO model to support multi-line item optimization back in 2000, actually supported this level of modelling back in 2000, an average of five-plus years before the majority of SSDO solution providers did