In the past year, we saw a backlash against the emerging consensus that Autonomous Procurement will be a key part of the future for procurement. Although it’s been a year since Andrew Bartels, a VP and principal analyst with Forrester and former guest speaker at a Keelvar Konnect event in New York City, launched a broadside claiming ‘Autonomous Procurement is a Really Dumb Idea’ (December 2019), the topic in the past few months resurfaced as a discussion on social media. Was this a pragmatic critique of a misplaced gambit by technology companies, or was it a nostalgic appeal to reassure personnel in the field that their positions are safe?
Let’s understand the argument that was made and the context. Andrew Bartels was speaking after briefings with several suite vendors who were extolling a vision of Autonomous Procurement in which human intervention was not required. “The software handles most of the tasks of purchasing, from deciding what to buy and who to buy it from to making the purchases without human involvement.” He goes on to say that this is a really bad idea. The rationale cited for this is that:
Let’s assess these four points in turn. The implication in the first reason is that the only place it makes sense is in simple inventories for direct materials. This ignores applications such as intelligent supplier discovery, spend analytics or any tactical sourcing activities in indirect categories and logistics.
At Keelvar, we have witnessed organizations with hundreds of personnel running routine auctions or RfQs and invariably executing them suboptimally with poor choices for bidders or limited feedback to drive competitive tension. Sourcing excellence requires precision and attention to detail that is difficult to scale via human-operated processes. Several companies, such as Siemens and Coca-Cola, have already publicly testified to the success of Sourcing Automation across various spend types. More success stories will be publicly told in the coming months and years.
One might argue that Bartels could be justified if the second point was true: namely, that automation overpromises and will under-deliver. Keelvar has been deploying our e-sourcing automation solutions over the last 18 months, and without exception, each of those customers is scaling up their adoption of those solutions and reporting impressive results. Siemens, for example has reported automation has saved their buyers up to 93% of tactical work time.
I find the third and fourth claims are interrelated, positing that autonomous procurement won’t find adoption because it scares people and is hostile to their jobs. This is an argument that can easily be dismissed just by looking at how technology advancements have transformed other industries. No doubt electricity scared wax candle producers, and satellite navigation on ships scared lighthouse keepers. Just as the agricultural revolution destroyed some jobs, it created many more. From tractor manufacturing to crop protection technology, people adapted, and ultimately their wealth increased as society became more efficient.
The real threat to employees is trying to maintain the status quo or an inclination to believe a comforting message trying to reassure you that there’s nothing to see here and don’t fret about autonomous procurement. If companies don’t embrace technological progress, then their competitors may do so before them, and that is ultimately Darwinism in action. Companies should be looking at new opportunities this opens up for them.
Andrew Bartels was somewhat justified in challenging suite vendors for the lack of intelligent systems in their offerings. Suites are large and broad product offerings built on outdated architectures that are ill-equipped to tackle the data engineering and data science challenges of this decade. Younger, best-of-breed solution vendors are advancing technology at a faster pace than established suite vendors, so drawing conclusions based on discussions with the latter will yield different take-aways. This is the wrong place to search for innovation because these are older vendors with technology stacks that limit their flexibility to harness more modern methods.
To get a more accurate, holistic perspective of how autonomous systems will actually impact various areas of procurement, analysts should be at least equally focused on the disruptors, the upstarts and innovators whose livelihoods depend on delivering truly transformative solutions. For those providers like Keelvar and their customers, the future is already here and impacting the real world, just not equally distributed.
Bartels has a point in that he may wish to see autonomous systems being demonstrably proven in a public setting. Let’s get transparent, shall we? That is why Keelvar developed the Human+Machine Contest for sourcing; to assess whether a fully automated, AI-driven sourcing bot can compete against an expert human buyer. Similar to Garry Kasparov vs. IBM’s Deep Blue, we will find out whether technology has exceeded human performance when it comes to routine, tactical e-sourcing task work.
As chess was conquered by machines, so too will manual processes in procurement be automated. Autonomous sourcing, and procurement more generally, will be a long game of chess but ultimately a checkmate position is approaching for most of the current ways of working. But this should not be confused with being pessimistic. On the contrary, new roles will emerge and people in procurement will find ways to flourish by exploiting new areas of value creation. There are ample tasks that humans are better at doing and they will have more time to focus on these. These roles will be more fulfilling and strategically challenging, and the world will keep spinning with Human and Machine working in tandem together.
Join Keelvar at the Human+Machine contest with a guest appearance from Garry Kasparov in 2021. You can also watch on-demand our special Contest Prequel that aired on December 3rd. https://contest.keelvar.com