Sourcing Optimization Should Be Accessible, Flexible: The Case for Rethinking a Misapplied Technology

Sourcing optimization solutions arose, in the simplest terms, to help procurement organizations find optimal solutions to their most complex sourcing challenges. Faced with balancing overlapping and sometimes conflicting constraints — including supplier location, item coverage, cost of sale, switching costs and risk — sourcing professionals found that more traditional, manual paths to decision-making were unable to account for all of these varying factors.

This is the problem that optimization solutions aim to tackle.

By adding numerous constraints for business rules and costs into the sourcing equation, these tools are able to quickly parse large amounts of data and assess multiple scenarios, saving analysis time, negotiation time and, of course, money.

But sourcing optimization tools also have a persistent shadow: a high barrier to entry due to low usability. Naturally, the companies that most often tackle complex sourcing challenges are those with global, dynamic supply chains.

First-generation sourcing optimization providers had lofty goals, focused on solving only the most intricate problems for the most advanced procurement organizations. Yet this mindset has limited procurement’s perception of what sourcing optimization can be. While these high-powered tools can accomplish amazing analytical feats, the effort and sophistication required to reach a desirable outcome is extensive. It should be possible to apply these capabilities to numerous additional scenarios — without the high technological barrier to entry that users have come to expect.

Said another way, sourcing optimization can (and should) be applied more broadly to a variety of sourcing scenarios, bringing a simpler approach and guided exploration — designed with the everyday sourcing user in mind. With the value it can bring to categories and customers alike, why should sourcing optimization remain out of reach beyond the most complex logistics bids?

To understand how this situation arose, it helps to consider the rationale behind how first-generation sourcing optimization solutions were designed.

Tools Built by PhDs, for PhDs

One of the major reasons sourcing optimization has failed to become a mainstream tool for procurement organizations is the considerable training, customization and sophisticated coding required.

Today, the target user for one of these sourcing optimization tools is someone with advanced mathematical, technical and domain expertise (e.g., in transportation sourcing) — hardly a common skillset. Unfortunately, this means the benefits of the technology are only accessible to a small subset of the market.

This is why most sourcing optimization tools are limited to scenarios in logistics and freight sourcing. With myriad requirements such as multiple modes of transport (lanes), tariffs, lead times and more to consider, identifying the ideal logistics strategy is far from easy. Thus an effective sourcing optimization needed to be able to apply numerous hard and soft constraints, compare bids across thousands of lanes, and compare various award splits effectively. But as a result of this category specialization, these platforms were not built to address any other scenarios “out of the box.”

Other industries that could benefit from optimization capabilities — like food and beverage, pharma and biotech, and manufacturing — have seen their needs go largely unmet.

Consider, for example, the decisions a manufacturer now grapples with, especially amid globalization:

• Should a product be produced off-shore or locally?
• How should various costs (e.g., raw material costs, tariffs, transport costs) be weighted relative to one another?
• How can we represent the interdependent nature of these costs in the optimization?
• Taking a step back, should the company manufacture this product at all, or is it simpler to contract a third party?

Decisions like these spin a complex web of conditional logic, which is much easier to task to a sourcing platform than to a human analyst. The demand from the rest of the market exists, but most optimization tools still have not adapted to the needs of other industries.

The Missing Factor: The Human Touch

There’s more to solving this problem, though: A great category manager does not want the tool to make the decision for her. Instead, savvy sourcing pros are looking for a guiding hand, a set of potential solutions to be tweaked further.

It is simply not worth the effort to have an algorithm analyze the most minute factors. Indeed, solution providers are never going to convert every mental sourcing rule into a compute rule. Not all of the nuances that drive sourcing decisions make sense to operationalize, so some of them should remain with the experience and expertise of the individual team members.

Sourcing professionals have a different thought process for every supplier, category, item and event. This means that each sourcing event across categories is unique, so asking a user to code every constraint needed for a particular model is inefficient. For a decision so specific (and perhaps arbitrary) as setting a $5,000 threshold to continue with an incumbent supplier rather than switching, the value of standardizing it as a rule is questionable.

Instead of serving every myopic use case, the scope of sourcing optimization should instead cater to the average sourcing user, helping her get close to the best decision while allowing her own expertise to fill in the blanks along the way.

To make that happen, providers must put themselves in the shoes of the everyday user and realize only a select few can take advantage of a high-end linear optimization and stats package. Every organization can realize significant benefits from optimization, but the poor usability of the current generation of solutions has poisoned the well.

To reset the narrative, solution providers need to design optimization tools to do what sourcing professionals need most: narrow down the options and let individuals fine-tune the final decision.

A Better Path: Guide Decision-Makers to Better Decisions

The reality is that sourcing decisions depend on too many factors to be completely quantified or churned out by a computer. We can build software that gets us most of the way there, but it should empower the decision-maker rather than do his or her job outright.

Accomplishing this requires a change in priorities: delivering flexibility and insight rather than the complexity and customization required by previous-generation solutions.

Much like a vintage radio, the process now finds a balance between machine and human. Optimization serves the role of the large tuning knob, while the smaller knob requires the deft touch of a human to fine-tune the perfect result.

What does this look like in practice? Next month, in Part 2 of this series, we’ll explore just what the future of sourcing optimization holds: a smart, approachable tool that brings all of the benefits of old approaches — without the headaches.

This Brand Studio article was written for Scout RFP, not as Spend Matters editorial or analyst content.