Guest Blog Rich Wilson

Part 2: Looking back and moving forward: The past, present and future of e-sourcing

September 11, 2019

author:

Part 2: Looking back and moving forward: The past, present and future of e-sourcing

In this two-part blog series, Dr Rich Wilson outlines the rapid acceleration of e-sourcing since its inception in 1990: the journey it has taken, the highs and lows that exist today, and how this will influence tomorrow’s best practice.

Dr Rich Wilson outlines the history of e-Sourcing, and the processes it has shaped as we know them today in Part 1.

No doubt, current processes and technology signify that it is time for another technology revolution in e-sourcing today. We don’t have to look far to know where that revolution is likely to originate. Artificial intelligence promises not only to address aging technology tools but also the paucity of skilled sourcing professionals able to use them. Sourcing Automation is the technology revolution we need now.

Interestingly, while many of the best practices in e-sourcing have become clear in three decades of its development, thinking about sourcing automation remains cloudy.  I will point to a few ideas one encounters in discussions about AI and automation which I feel point in the wrong direction.

The most significant opportunity in sourcing automation is not to automate the way we do things today. We need to automate the way we could do things today, in a better way. That is, to automate best practice. Why bother to automate processes that often don’t deliver the outcomes we would hope to achieve, when we can design an automated process that achieves something better? 

From a practical point of view, I think it is important to understand that these are the early years of this next sourcing revolution. There is an opportunity now to do things in an incremental fashion. 

For a long time in the e-sourcing world, people were presented with the idea that they had to adopt an all-encompassing suite of solutions that did everything.  Now, pioneer adopters of sourcing automation do not hesitate to start small. Incrementalism is not a mistake in adopting an automation strategy as it allows for the learning which we know is important for success in the long term.  For those ready to take a leap, however small, into the new world, Keelvar should be the first stop. 

Keelvar was established in 2012, and began work initially, not on sourcing automation, but on sourcing optimization. You don’t invent these capabilities overnight. The number of companies that have blithely claimed that they were going to produce a fully functioning combinatorial sourcing optimization application is legion, but one can count those who succeeded on the fingers of one hand.  An underestimation of the required effort and the discipline to refine an optimization offering was the rule, not the exception.

I joined the board of Keelvar because I was deeply impressed with the vision of its founder, Alan Holland, and the remarkable team of AI experts he has assembled. There are only a handful of technology providers today who have even a few of the essential ingredients for developing and delivering sourcing automation.  And among these, Keelvar’s capabilities are the most complete.

What then are the essential capabilities of a true Sourcing Automation provider?  

First, without question is a solid foundation in sourcing optimization. Combinatorial optimization is a precondition for taking the next step in sourcing automation, because it is the only proven technology for complex decision making with dozens, hundreds, even thousands of business rules and which is robust in the face of unpredictable inputs. This is not achievable with Robotic Process Automation; complex procurement decisions can’t be automated by rote learning. Sourcing optimization is not automation, but it’s a precondition for it.  

Keelvar has been in the business of delivering sourcing optimization for seven years. Of all the providers today with sourcing optimization capabilities, it has the most recently developed and current technical base with the longest expected life span.  

As a spin-off of one of the world’s leading AI labs, however, Keelvar’s vision from the start has been to deliver Sourcing Automation.  With all the talk about AI today, it is easy to miss that real expertise in the field is still quite rare. AI expertise is not something that one acquires readily in the course of a college education. Understanding AI theory and methods and mastering the tools necessary to design and build AI applications is not commonplace. And it’s not likely to be commonplace for some time to come. Keelvar has it now and has had from the very start. This, I believe, is the second key requirement to be a true sourcing automation provider.

Finally, it is vital to have an AI enabled technology stack, the kind that enables Keelvar to incorporate Google’s TensorFlow in its Sourcing Bots and to leverage other third-party AI tools.  This is required for a sourcing automation provider to maintain an accelerated pace of development. No successful sourcing automation provider of the future can afford to be an island.  And this, I think, is a third essential capability which sets Keelvar apart today.

A few months ago, McKinsey consultants went through an exercise to estimate the value creation potential of AI in various aspects of business.  One area that stood out from all the others was supply chain where the return on investment is projected to be as much as 1.3 trillion dollars in the next twenty years. 

Given this potential, the relatively scarcity of expertise in this area, and the number of firms that have what it takes to deliver real sourcing automation, first mover advantages will prove significant. 

Procurement leaders should think seriously about getting started soon. Remember: start small, start sustainably. Incrementalism is not a mistake for adopting a long term strategy.

Ready to learn more? Register now for Keelvar Konnect ’19 Zurich, November 6th.

Dr Rich Wilson is a non-executive Director at Keelvar. He is the former CEO of CombineNet, and founded and led operations research at USAir. Rich also held the role of transportation consultant at Oracle. Today he enjoys offers classes in technological and social change in Pittsburgh, Pennsylvania.