Procurement Expert by Gemma Howard-Sandy

The robots are coming. Except this time, they really are!

Part 2

Continued from Part 1

What's Marketing Got to do with Procurement Anyway?

Naturally, an established aim for procurement is cost saving. But managing total value beyond simple economics is becoming increasingly significant. Total value management includes risk management, as mentioned in part 1, but more and more, it encompasses brand management too.

Traditionally, brand management is something procurement professionals have left to the marketing department or to external marketing agencies. However, an organisation’s brand is not just about selling goods and services to the client; it’s also about trust and current expectations/ standards. Procurement personnel are able to help create goodwill by engaging a supply chain that takes ethics and sustainability seriously.

Corporate social responsibility (CSR), modern slavery concerns and the gender pay gap issues are all "opportunities" that procurement and marketing can realise together. But will AI be able to keep up with the ever changing landscape affecting businesses in the 21st century?

AI, RPA and Machine Learning From a Human Perspective

Could procurement RPA evolve to guiding people to the correct purchase? Bertrand clarifies, "No, not RPA, maybe AI via the approach of an assistant. A Procurement assistant is deployed to handle demands from the rest of the organisation in order to replace or "augment" traditional eProcurement solutions. Requesters interact with a bot that proposes solutions based on:

  • the needs identified during the conversation
  • the Procurement strategy (preferred suppliers, preferred items, contracts in place)
  • other factors (purchasing history, real-time availability of products, context, etc.)

He continues, "The approval process also happens via chat. If available, the chatbot adds the approver to the conversation, creating a group chat. Or, the Procurement Assistant opens a new one-to-one conversation with the relevant approver. Approvers can then ask the chatbot how much of the budget is left and then immediately approve/decline the request without leaving the chat. The same can happen for other process steps (order confirmations, goods receipts etc.). The assistant initiates discussions to ensure the process is compliant and efficient."

Could AI provide summaries, recommendations and advice regarding supplier appraisals, supplier performance, risk management and industry regulatory compliance? Or is this simply a human being with the correct level of experience and/ or a qualification or several?

Automation is not the solution to a broken process. It would be far more beneficial to solve any process issues first and then look at automation if the business case still supports it.

A high-volume manual process is potentially a “no-brainer” for RPA consideration. A lower-volume manual process will require much more justification of course. It might still be an option if, for example, it allows your sales or service delivery team to deliver on your clients’/ customers’ expectations better or faster but the business case for it ultimately gets weaker, the lower the volume. Another “no brainer” scenario is if automation can help reduce or eliminate costly compliance mistakes, resulting from human error, within your industry. As a side note, one of our clients has implemented this type of RPA across their healthcare decontamination and sterilisation business and it allowed them to take on 30% more work; what could 30% more work do for your business?

Software developers, working on AI and RPA, will most certainly need the right analytical and data science skills to understand the opportunities and limitations of machine learning in procurement. Bertrand enlightens, “And, foremost, the right data to train the machines. Today, the highest costs to implement/ build AI is not about the algorithm. Many open source frameworks are available for machine learning and NPL and are applicable across industries. However, datasets to train these algorithms are specific to a company/ process/ context and due to the vast amount needed to train AI, this represents a higher cost. AI will not (yet) auto-magically learn. Most AI is supervised by humans that train/ correct the machine.”

From spend analytics and contract analytics to risk management and supply chain knowledge, the developers working to achieve the potential will need a diverse set of skills that may be hard to come by. The skills sought now for an effective procurement professional differ from 10-15 years ago but do procurement professionals and software developers have enough mutual ground for RPA to reach its potential?

Many professionals enter procurement to be hands-on problem-solvers across the business. From finding funds for new projects, products or growth, to negotiating with and optimising a robust supplier base. With AI doing the rest, they can certainly get on with it but with the more process driven tasks currently used as a breeding and nurturing ground, for less senior roles, could this create problems for procurement in 10-15 years’ time?

Human Judgement, Ethics and Morals

RPA clearly works for relatively simple, if/ then type rules. For example, "if a customer number in file X matches a SAP customer number, then update SAP customer file." However, if human judgement is required will AI/ RPA ever be able to replicate this without presenting the bias or prejudice of the programmer? Bertrand explains, "or the bias (or poor quality) in the data used to train it… or the trainer, such as Tay, which was an AI chatbot, released by Microsoft in 2016."

This is the area of AI/ RPA that I’m most fascinated by and also most confused by.

It was Emerson M. Pugh who said,

“If the Human Brain Were So Simple That We Could Understand It, We Would Be So Simple That We Couldn’t”

We’re complex and diverse in our understanding and initiation of tasks. Our execution will always present different results and wins because of how our minds work. Humans also care and have an ethical decision making process based on fact. How can any of this be replicated when scientists don’t fully understand all of these functions and outputs in our own brains?

Bertrand concluded, "Very true, however, not impossible (and that is very creepy)........."

Buzz Words, The Now and Back to the Past

Are we going to look back at this in 10, 20 or 37 years’ time and reminisce that none of this materialised?

Procurement is changing. It is now a multifaceted proposition that demands that professionals focus not only on the traditional procurement activity of cost management, but that they take a broader, total-value outlook. With the ability to relegate many administrative processes to robots, organisations now demand that their procurement personnel undertake more value-enhancement activities, such as actively managing the supply chain, managing risks better and helping with brand building.

The technology is still in an early growth stage, but it holds immense potential: dramatic improvements in accuracy and speed for some functions and potential cost savings compared with more labour-intensive processes. The robots may or may not be coming, but what is for certain is that the 7 most expensive words in business are; "We have always done it that way."...............

Bertrand's Final Thoughts:

"Without allowing our 'humanness' into the procurement process, I think we will be too focused on technical procedures. In order to be drivers of change we must see the similarities and differences between what we do when we shop for ourselves and when we are simply following the procedures outlined by CPOs."

"I believe that Procurement (and any other function) is facing new challenges introduced by new technologies and it has to boldly go where no man has gone before. And, to do so, it has to find the right balance between its human and its Vulcan sides. In other words, find the way to use new technologies that will enhance Procurement people."

"I am a true believer of “people first”, especially in Procurement. I also believe in "people + technology", by opposition to "people vs. technology". And, this is an area where Procurement still has a long way to go as it must be more technology savvy."

"Also, new technologies are already here. So, it is important to understand what they are and what they bring, to define how to get the best out of it. This is important for the profession and this is important for the next generations of workers. No doubt that automation is changing work. And the thing is: it is just the beginning!"

"Without going into too many details, the future of Procurement work can be described as “Cognitive Procurement”. It can be defined as:

  • Using systems and approaches that are able to learn behaviour through education
  • Managing structured and unstructured data
  • Supporting forms of expression that are more natural for human interaction
  • Continuing to evolve as computers experience new information, new scenarios and new responses
  • Unlocking new insights and enabling optimised outcomes"

"Some of the applications illustrating the "tech. + people" (or Human + Vulcan) idea revolve around the concept of an assistant. The role of the technology being to guide and recommend and the final decisions being left to people."

"Today, the highest costs to implement/ build AI is not about the algorithm. Many open source frameworks are available for machine learning and NPL and are applicable across industries. However, datasets to train these algorithms are specific to a company/ process/ context and due to the vast amount needed to train AI, this represents a higher cost. AI will not (yet) auto-magically learn. Most AI is supervised by humans that train/ correct the machine" - Bertrand Maltaverne

Bertrand Maltaverne

Procurement Digitalist

Bertrand Maltaverne has extensive experience in the area of procurement, and more precisely, in the impact of technology on procurement processes and organisations. He works at Ivalua as a Solutions Consultant, helping organisations achieve success in their digital transformation of their procurement practice. In previous roles he's focused on procurement technologies and processes in both direct and indirect procurement. In parallel to his professional career, he is active on various social media, and he also blogs.

Bertrand Maltaverne