“In my digital supply chain team, I have no procurement people.”
Virginie Vast, head of cognitive procurement and digital sourcing at Vodafone, described her unconventional setup to an audience of senior professionals at the 2017 World Procurement Congress (WPC). The line was delivered with an almost daring self-confidence; a subtle yet knowing glint in the eye that said to her listeners: I have something you want and you and I both know it.
Vast’s words act as a lens, held up to the future of the procurement function. Her words tell CPOs that the emergence of artificial intelligence (AI), or cognitive technology, means the teams that are fit for today will not be the ones that are fit for tomorrow.
AI demands a new approach to skills in procurement. CPOs will themselves have to adapt, but not as fast as they will have to recreate their teams, because new people with new expertise are required in order to harness AI. Those executives leading the change are already creating procurement roles that would not have existed five years ago, for people from diverse backgrounds and different experience. Instead of buyers and category managers, Vast’s team includes data scientists, coders, robotics specialists and even psychologists.
The rate at which AI is becoming more sophisticated makes it difficult to ignore fears that computers will substitute some people. Yet senior procurement figures who are familiar with the technology say the point of AI, for now at least, is not to replace workers, but to enhance their roles – a position that some experts share.
“If it [AI] marginalises or displaces some of these low-level administrative and data-gathering operational roles, that’s probably work we don’t want to do anyway,” argued Dan Carrell, vice president of global procurement, IBM, in an interview with Procurement Leaders.
Cuts to the team may be made as some ‘traditional’ procurement staff are no longer needed, but it is believed most people will remain. The question is: what will they be needed for? Procurement executives hope that AI will give their staff the time and space required to grow into data-driven strategic managers, focused on stakeholder alignment, innovation and product development.
But whether dreaming up new roles or reimagining old ones, the emergence of AI compels CPOs to address a fundamental question: what people and skills does the procurement function need to succeed? Technological change is fast and fickle; if executives are not working out their answers to this question now, they may already be behind.
“We need completely new profiles in procurement,” Henrik Larsen, CPO, Maersk, tells Monitor. “We need data scientists, physicists and even librarians.”
Not so long ago, data science and computer programming would have been thought of by most CPOs as immaterial to their objectives; potentially interesting, albeit remote, areas of work that were left to colleagues in IT. Today, however, procurement leaders at the forefront of technological transformation regard individuals with data science and programming skills as essential to their digital ambitions. AI, for one, will be difficult to make a success of without them.
The technology uses computational techniques such as natural language processing and classification modelling to interpret huge amounts of structured and unstructured data – more than is feasible for any human – and establish more patterns and more connections. But, today, an AI system cannot be fed any old information and be expected to provide useful answers in return.
AI requires training. For that, “you’ve got to have a baseline of [good quality] data,” said IBM’s Carrell, which means hiring people who know “how to look at data…how to put it together…how to extract insights.”
That’s where people such as Edward O’Donnell, global procurement data officer at IBM, come in. O’Donnell personifies the new approach being taken to skills within procurement. A veteran of the profession, O’Donnell has seen a lot in his more-than-30-year career. Yet time appears to have done nothing to dull his enthusiasm for new ideas, especially those relating to AI and its potential to enhance the work procurement does.
“Today is a great day,” he tells Monitor from his office in New York. “I’m excited because I’m meeting later with IBM executives to demonstrate a robot, or, as we call it, a cognitive intelligent agent, that my team is developing. We’re taking historical transaction data and collating it with market sentiment fed from Watson content analytics, so the robot advises our procurement staff on what they should be buying, starting with those in the labour and services category, and expanding from there.”
A chief data officer? In procurement? Developing robots? IBM’s progression seems a world away from what most would consider within the boundaries of the function. Yet O’Donnell is adamant that the direction his function is moving is where most CPOs want to go, “they just don’t know how to get there,” he says.
The scale of the changes being made at IBM demands thoughtful planning. The teams, people and skills required to drive the change occurring at the company are all different to what they once were.
We have brought data scientists into the function because we need them to be embedded within the procurement agenda
“I have a team of about 30,” O’Donnell says. “Some data scientists, doing master data management, creating models and algorithms, writing code for analysing unstructured data and turning it into structured values.” For these purposes alone, working at IBM has its perks. “Of course, we get to cheat a little bit,” he says mischievously, referring to procurement’s ability to draw on IBM’s technological expertise. “We work with our colleagues in Global Business Solutions and the Watson development team who are right on our doorstep and do this for a living.”
Not every procurement function has direct access to the team behind a world-beating supercomputer nor the budget to create a data office all for itself. However, that does not preclude a change in the approach taken to skills. While speaking at WPC, Vodafone’s Vast urged attendees to rethink the skills required to support their technological ambitions, as well as where and how this expertise can be acquired.
If CPOs are not looking at ways to reshape their teams, if they think the changes being made by Larsen, O’Donnell, Vast and other forward-thinkers are not for them, they may soon find themselves chasing the early movers.
Like O’Donnell and his team at IBM, procurement chiefs can probably borrow expertise from adjacent business functions if required. This approach was initially adopted by Vast, who sought assistance from Vodafone’s business intelligence unit, in which data scientists were already working. But what her function shares with others pushing the margins of technological change is a recognition that experts in data, analytics and computing must feature in procurement’s long-term plans.
Leaning on other departments in the organisation has its limitations. For Maersk’s Larsen, bringing new people into his function is about ensuring they have the necessary time and scope to focus solely on contributing to procurement goals. “We have brought data scientists into the function because we need them to be embedded within the procurement agenda,” he says.
This approach is more than symbolic. As O’Donnell points out, there are tangible benefits to bringing new skills into the function. “Not having to pull on our IT staff all the time makes us nimbler, more agile,” he says. O’Donnell’s role as procurement data officer is only about 18 months old, yet already he has seen “the pace of deployments in analytics and cognitive capabilities begin to pick up,” because of his team’s work. “We’re in the early stages of adoption, but we expect these tools will help save us $50m in 2017 and even more in 2018,” he adds.
The need for new skills places the onus on procurement chiefs to look for talent in new places. Not everybody in the function will be suited to a new role, as O’Donnell was; some staff will need to be hired in. To attract the kind of new people she wants, Vodafone’s Vast believes partnering with schools and universities that run leading AI and robotics programmes is essential. But, she also recognises that the next technology whizz may exist within the team. Procurement executives, she said, must understand their teams’ existing capabilities, for who better to embed within procurement than somebody who already knows the function and has the desired expertise?
HR may be able to help a CPO single out their digital superstars, but open communication with staff is as good a place as any to start. Leaders of the function need the right mindset as much as anything else. “If you have the skills, tell us,” is the message that Vast sends her people via social media – the digital purist that she is. “I am amazed by the skills in my team that have nothing to do with supply chain, but are critical to our digital transformation,” she told the audience at WPC.
If you find yourself at the reception in Vodafone’s Luxembourg offices, you’ll be greeted by Pepper, a companion robot created by Japanese tech firm SoftBank Robotics, with a friendly “hello”. Vast said that Vodafone was paying the company that supplied Pepper to write the code that enabled the robot to do this, until one of her team changed the greeting on 14 February to “happy valentine’s day”; all they had to do was connect to the wireless internet and rewrite the code, she said.
Procurement leadership, according to Vast, “really has to rethink what people are capable of,” because the emergence of AI makes the acquisition of new skills essential. As far as traditional procurement staff are concerned, some old skills may even become obsolete. Not everybody in the team can be a data scientist or a coder, yet those who face aspects of their job being dislodged by the introduction of AI do not escape the need to adapt to the changes ahead.
The question procurement executives should always be asking is how they can buy better, faster and smarter. By reducing the amount of time spent sifting through information about suppliers, contracts, expenditure and risks, AI is expected to help the function answer that question more easily. But if CPOs fail to recast the role of the traditional procurement professional in such a way that complements their use of AI, the function could remain locked in the same tired cycles.
Year after year, Procurement Leaders’ research into the priorities of procurement chiefs reveals a desire to assign a greater number of people to higher-value activities, such as stakeholder alignment, supplier relationship management and innovation – but it largely remains a desire. “Collaborate tomorrow; negotiate on price today,” Jonathan Webb, author of CPO Planning Guide 2017, writes of the impression left by findings from the study.
Those in procurement who are familiar with AI believe it represents the catalyst for change that the function has been missing. “I like to think of AI…as an intelligent agent that allows us to bring new value to the conversation. It enables us to elevate what we do, the value we bring and the outcomes we deliver,” explained IBM’s Carrell.
“I don’t want to use the phrase ‘super-buyer,’” says O’Donnell of the future procurement manager, assisted by AI. “We used to say that years ago. But, when you look at what cognitive technology can do, at the kind of information it can give buyers at their fingertips, it completely changes the experience when we engage suppliers, stakeholders and clients.”
The speed at which AI systems can produce that information is bewildering.
O’Donnell explains how comprehensive supplier performance reports, including more than 40 benchmarked financial performance metrics, which he estimates would take a week for a buyer to prepare, can be compiled by a cognitive computer in a matter of minutes and viewed on-demand.
The potential of AI is clearly great, but to capitalise on it, ‘traditional’ procurement staff must demonstrate new skills. For CPOs, two priorities relating to the reeducation of people whose work is displaced by AI are beginning to emerge. First, at least some must be proficient in their use of the technology and the data to which it gives access.
Scientists, coders and statisticians may be brought into the function to help exhaust the possibilities AI presents, but why limit data-driven thinking to a small number of people in the team? “When I do interviews now, one of the first questions I ask is: ‘How do you manage data?’” says IBM’s O’Donnell.
The importance of data is such that O’Donnell expects a basic level of know-how to exist right across the team. For this, he has the “data wrangler” programme – a training initiative designed to teach category strategists some data-science basics and bring them up to speed with the cognitive technologies available to them. Within IBM’s internal procurement academy, O’Donnell has also created what he calls the “analytics and cognitive corner”, through which video tutorials and other educational content pieces related to the different aspects of Watson’s procurement applications can be accessed, in order to raise levels of awareness and adoption.
“One significant change we’ve made is the introduction of data scientists, but another is the development of this broad, general programme to upskill our procurement staff,” O’Donnell explains the way he has attempted to instil a data-driven ethos across the function.
Forward-thinking CPOs I speak to recognise that cognitive technology will free their functions from the time-consuming, process-based, transactional elements of procurement; they see an opportunity to step into strategic business partnering
Data should not be the only focus of efforts to upskill staff displaced by AI, though. Second, with Carrell’s “intelligent agents” at their sides, buyers’ ability to communicate effectively, empathise and influence only grows in importance, as these human aspects cannot yet be appropriated by computers. It is on these foundations that procurement’s future as a strategic function is based, according to Lucy Harding, partner and head of global practice, procurement and supply chain at recruitment firm Odgers Berndtson.
Harding recently sat down for lunch with 12 CPOs to discuss the future of the function. She tells Monitor that while technology and digitisation were high on the agenda, they were seen as a means to an end. Future CPOs, Harding argues, may well be among the most technologically savvy in the business, “but with the caveat that they will still need to have all the qualities of a broad-minded strategic leader”.
“Forward-thinking CPOs I speak to,” Harding adds, “recognise that cognitive technology will free their functions from the time-consuming, process-based, transactional elements of procurement; they see an opportunity to step into strategic business partnering.”
Maersk’s Larsen agrees that “the collaborative agenda is becoming much more important” to procurement’s role. For his strategic managers, Larsen advocates a two-one-two model by which they should divide their time: two days with stakeholders, one day at their desks and two days with suppliers. Assigning buyers to higher-value activities means addressing questions relating to stakeholders’ priorities and suppliers’ capabilities – questions that can only be answered by spending sufficient time talking to the right people.
AI is thought to afford managers more of that time, but, as machines take on more work, CPOs are forced to grapple with fundamental questions about the people in their teams. Conferences and boardrooms echo with talk of data science, code and web development as executives think excitedly about ways to thrive in the age of AI. But as the technology spreads, the procurement function’s bread-and-butter roles will need to be reconsidered as much as new responsibilities conjured up. Leading procurement chiefs are already defying norms by rethinking job roles, activities, people and skills. Others must begin to do the same, for the teams that thrived in the age of unintelligent machines will not be the ones that do so in the age of AI.
Opinion is divided on the risks artificial intelligence (AI) poses to people’s livelihoods. Elon Musk, CEO, Tesla, has said that AI represents mankind’s greatest existential threat. Some experts, meanwhile, decry the idea that a computer could somehow appropriate the irrationalities of human emotion, value and relationships. Yet Jurgen Smidhuber, codirector of the Dalle Molle Institute for Artificial Intelligence Research in Switzerland, paints a different picture.
AI technology is “completely stupid” to begin with, he explained during a presentation at the World Procurement Congress 2017, but feeding it with data enables the technology to learn through pattern recognition to imitate experts and, eventually, compete with them. The potential of the technology is “superhuman”, according to Smidhuber, as it is able to learn far more from data than a human could. It is easy to see Smidhuber’s point in the context of a global procurement function.
A category manager, for example, can only be in one supplier meeting at a time. A cognitive machine, however, can learn from data captured during every supplier meeting ever held by all category managers in an organisation. Does that mean the machine replaces the category manager? Not yet. AI is being used to crunch massive amounts of data to support procurement decisions, not make them. Will that change in future? Possibly.
Nevertheless, Lucy Harding, partner and global head of practice, procurement and supply chain at recruitment firm Odgers Berndtson, says procurement chiefs in her network view AI with optimism – they see it as something that will help them achieve their strategic goals.
Procurement chiefs should not take the rise of the technology lightly, although not for the reasons you might think. Those leading the trend recognise that the most urgent threat is not the loss of jobs to machines, but the loss of standing in their businesses to their own failure to prepare their teams for the changes that AI portends.
Edward O’Donnell’s role as global procurement data officer at IBM is an unusual one. But does his appointment cast a lens on the future of the function? O’Donnell’s overarching responsibilities are two-fold: he looks after procurement’s data and IT strategy while ensuring it remains aligned with that of the enterprise; and he is required to seed the function with people who have the skills required to take advantage of data, analytics and cognitive technologies.
Matt Holzapfel, data evangelist at Tamr, whose products use machine learning to clean and prepare data for analysis, works closely with his company’s customers. Holzapfel’s experience tells him that CPOs who “drive the IT roadmap” with the help of data, analytics and technology experts in their teams are the ones who fare best in the transition to cognitive technology. Regardless of whether a procurement data officer is brought in to put these elements in place, doing so must be somebody’s responsibility. Increasingly, that somebody is likely to sit in the procurement function.
When technology is changing procurement’s work in original ways, people must be able to acquire new skills. In all ranks of the function, learning itself is an increasingly attractive quality in people. “The ability to lead the procurement function through changes that advancements in technology bring is vital,” says Lucy Harding, partner and global head of practice, procurement and supply chain, Odgers Berndtson. Top CPOs, she adds, display intellectual curiosity, learning agility and are resilient in the face of change.
But what about their teams? Procurement chiefs may lead through change but their staff are often the ones to execute it, meaning they must also be passionate about learning and problem-solving. It is not yet known whether curiosity is a trait that can be taught; academics have grappled with the idea for decades. Today’s procurement functions are seeing if learning can be inspired, too.
Clive Rees, CPO at Fujitsu, has been tracking staff performance for years. Recently, his team implemented new HR technology that uses dashboards to track people’s willingness to sign up for new training courses. The system alone will not distinguish the good learners from the bad, but it will give management a good idea of their team’s thirst for knowledge. In addition, “it can encourage people to go on these courses,” said Rees in an interview with Procurement Leaders.
Procurement’s attraction to AI currently lies in the technology’s ability to race through data at dizzying speeds, as Maersk discovered with the introduction of a process automation robot it calls Holger in 2016. The system is responsible for checking, validating and storing data in such a way that makes it readily available for analytics. The robot’s most eye-catching feature is its ability to analyse 150,000 rows of Excel data in 12 minutes, compared to the four or five hours it would take someone in the team to do the same job.
“Cognitive technologies allow us to move people in our teams up the value chain,” says Maersk CPO Henrik Larsen. “People are taken away from scripted process work, so they have time to focus on the real value-adding activities.” With the time and greatly improved access to data that Holger gives them, Larsen’s people can provide better support to stakeholders and think critically about process improvements. His team’s focus is moving from execution to strategy.
When procurement technology turns cognitive, what becomes of the CPO? New people, new roles, new skills – procurement chiefs’ teams are changing around them, but who are they required to be: Data and tech masters? Network controllers? People leaders? Business collaborators? The answer lies somewhere between all four.
“Artificial intelligence (AI) will change the shape of the procurement function,” argues Lucy Harding, partner and global head of practice, procurement and supply chain, Odgers Berndtson. “There is a growing need for people at junior and mid-manager levels with expertise in data, technology and analytics,” she says.
Can the same be said for procurement leadership? Not according to Harding. “My clients are not asking for deep technological expertise,” she explains, “they want leadership, strategic thinking, the ability to drive change.”
Henrik Larsen, CPO at Maersk, agrees. “CPOs do not need to be masters of cognitive technologies, but we must understand what they can do for our functions and our organisations. We don’t know this fully yet, so we need to maintain the network that will keep us up to date.”
“We also need to get our talent strategy right,” adds Larsen. “Diversity in the team is critical because while you still need people doing traditional procurement work, time must be freed up for the team to work on other areas too.”
No model of the future procurement function precludes the coordination of CPOs’ executive responsibilities with corporate ones. “The biggest criticism I hear of procurement functions and their leaders is that they’re not engaged with the business,” says Harding. Maersk’s Larsen says: “In the end, procurement will always be there to drive competitive advantage. Cognitive technologies are an enabler.”
This article is a piece of independent writing by a member of Procurement Leaders’ content team.