Autonomous procurement is no science fiction. It will happen. How can you expect it to change the nature of procurement as a discipline and a career path?
Nottingham, England, 1811: at a time when wages were being depressed to starvation levels and skilled artisans put out of work by the introduction of machinery operated by unskilled labor, weavers led by the mythical General Ned Ludd organised a campaign of smashing machinery. They became known as the Luddites. Ever since, the term Luddism has come to mean opposition to industrialisation, automation, computerisation, or new technologies in general.
More than two centuries of technological advances later, nobody smashes up machines. Because, unlike in 1811, automation is not destroying a way of life. Sophisticated levels of automation are accepted as the norm in manufacturing industry and other sectors as diverse as agriculture and finance. Generally speaking, the higher the level of automation, the higher the level of salaries. Manufacturing and process plants that use robots or other technology to automate routine tasks tend to have a highly skilled and well-educated workforce. And in future, many tasks will be performed, in part or in whole, autonomously.
Nevertheless, people are uneasy about rapid change and the insecurity it causes. And since the start of the first industrial revolution, the rate of change has accelerated. For the first 200 years, progress was mainly focused on gradual improvements in engineering and mechanisation and the harnessing of different sources of energy. Then came computers, then the Internet, and with them, digitisation. A whole new era. Even so, until now digitization has mainly involved the transfer of manual processes onto computers. Even now automation such as it exists is mainly limited to extremely low-level routine tasks where human activity is replaced by rules-driven robotic process automation.
Further acceleration is on its way with Industry 4.0, because of the sheer volume of data that can drive machine learning and the steadily increasing sophistication of artificial intelligence.
Autonomous production will rely on the harnessing of data and software to move from reactive artificial intelligence to prescriptive. For example, with reactive manufacturing, defects are discovered at the end of the line and the production team responds to correct the observed error. Until then, the factory keeps turning out defective goods. A prescriptive AI system, by contrast, identifies potential errors in advance and makes small changes to avoid future quality failures. These small corrective actions are made autonomously, in anticipation of defects, and thereby reduce the cost of non-quality.
We are now poised to see similar developments in procurement.
From automated procurement to autonomous procurement
How will this progress unfold? Spend Matters has identified four levels in a journey “that starts with technology that that assists buyers in completing tasks and ends with a platform that applies knowledge that is collected from buyers to do the tedious parts of their jobs for them”.
The four levels are:
Level One – Automation built on assistive intelligence
Level Two – Augmented procurement built on augmented intelligence
Level Three – Intelligent procurement built on cognitive intelligence
Level Four – Autonomous procurement built on autonomous intelligence
A truly autonomous procurement solution will not only have cognitive capabilities embedded throughout the platform but will build on those capabilities to automate entire sourcing and procurement processes without any buyer interference whatsoever, when the opportunity arises. Such a scenario is not imminent, but nor is it science fiction. It is something that we are moving towards gradually; our view of the destination is still rather hazy, but we can see it. With this ultimate state, systems will not only learn from humans and adapt their behaviour using cognitive abilities but also learn and adapt to new tasks and situations like an expert would, without always having to surface exceptions for human review.
Let’s put things in perspective
But even if it does not happen overnight, does this mean procurement professionals will ultimately lose their jobs? There is no short answer, but it is certain that many tedious human tasks and activities will be displaced. There are some tasks that robots and other technology are good at, and others that only humans can do. But let’s put things in perspective. According to a report published by the McKinsey Global Institute in 2017, only 5% of human occupations can be fully automated, although approximately 60% of occupations have at least 30% of technically automatable activities. The activities that it identified as most susceptible to automation are physical ones in highly structured and predictable environments, as well as data collection and processing.
Thus, up to two-thirds of jobs will change to a significant extent. Some jobs will disappear, but will be replaced by growth in other, more interesting activities. Check out this website. Enter “Procurement Clerk” and it will return a 98% probability that the job will eventually be replaced by what it calls “robots”, i.e. digital technology. But enter “Purchasing Manager” and the risk sinks to a virtually negligible 3%: the risk level is “totally safe”. With “Logistician” the risk is even lower, at 1.2%. It is easy to detect the pattern here: the more your job depends on human intellect and the less it depends on routine, the safer you are, even if aspects of your job can and will be automated.
In short, fewer boring activities, more value-adding and strategic activities.
I think there is a recognition that we cannot hold back technological progress, even if we want to, and that technological progress is an imperative in a dynamic, competitive environment. Nevertheless, fears persist. Some have expressed the fear that autonomous procurement will rob procurement professionals of critical activities such as sourcing, contracting, managing, and executing purchases with the suppliers of goods and services. But in my opinion, these are precisely the activities that cannot be handled by technology alone, with the exception of execution, and even here, there will be need for human intervention.
Autonomous procurement will use prescriptive AI to anticipate and correct small “quality defects” in execution before they happen. While each of these corrections is small in itself, they will add up to big benefits in terms of cost savings and performance improvements. But this will not replace the bigger decisions that will still depend on the ability of procurement professionals to make decisions based on a combination of data analysis and experience. In other words, procurement professionals will be freed up to focus on things like identifying new opportunities through category rationalisation. Once these activities are done, autonomous procurement will take care of formerly labour-intensive tasks such as setting up and executing an entire sourcing event from start to finish.
It is worth mentioning that increasing levels of automation in procurement will play a decisive role in attracting talent into the profession in future. Young talent and tedious, routine work do not mix well! As David McBride, Transformation & Strategy Director at real estate professional services company JLL put it, “Younger procurement professionals entering large organisations now expect to use cutting edge technology.”
In which areas of your role would you find autonomous procurement capabilities most useful? Let us know in the comments below.