Using AI Effectively (and Safely) in Procurement
July 24, 2023
July 24, 2023
Using AI Effectively (and Safely) in Procurement
The pressure on procurement teams to create strategic business value for their organizations has never been higher.
Industry advocates, such as Dawn Tiura, CEO and President of SIG (Sourcing Industry Group), believe that procurement needs a seat in the c-suite right beside the CEO, as procurement teams must be plugged into the deeper, longer-term planning and execution of an organization’s business goals. Procurement as a function has a crucial role to play in the broader effort to increase efficiency, improve process and employee experiences, and future-proof operating models.
But that’s… a lot. Plus, procurement still faces many of the same challenges it always has. Most modern procurement processes are complex, which is one reason few employees actually follow procurement processes as they’re designed. (Even if people know what the processes are, far too often the user experience is so onerous that people will work around and outside of them to get what they need. That results in low process adoption, which leads to waste, untracked spending, and additional risk.)
Accounting for the sheer number of moving parts in the supply chain remains incredibly difficult, as well. And there are always additional variables to contend with, like legacy databases and layers of approvals from disparate stakeholders.
The question, then, is how can procurement professionals go about achieving their lofty performance goals while at the same time finally addressing the challenges that are inherent to procurement as a practice?
One way is to embrace innovative technology. A great procurement team will use the latest technology, tools, and techniques to control spend, spot waste, develop key supplier relationships, eliminate redundancies, increase efficiency, identify various risks, avoid compliance issues—etc. More generally, they’ll use the resources and new capabilities made available to them to up-level their own capacity, and improve their organization’s performance from the inside-out.
And of course, few new technologies possess more potential to this end than new AI and automation platforms. High-performing procurement teams that are leveraging these tools are automating themselves out of tactical work and building an operational model that’s more agile, and scalable.
But AI is not a panacea. Here’s everything you need to know about how procurement teams should be leveraging AI safely and strategically in their work today.
AI in procurement: Getting smart about Cognitive Procurement
As in other fields, the potential of AI for procurement professionals is to help us work smarter. “A machine can think faster. It doesn’t forget. It can see patterns we can’t see. And so it’s very powerful when it’s built into a toolset correctly,” said Tiura, the CEO and President at SIG, in a recent episode of the Modern Business Operations podcast.
At a high level, using AI and machine learning to streamline, automate, and improve the P2P process is called Cognitive Procurement.
Cognitive Procurement up-levels procurement teams by helping them intelligently automate tactical and administrative tasks and allowing them to become, over time, more trusted advisors to the business.
A good example of an area of procurement wherein cognitive procurement can be especially useful is strategic sourcing, which involves observing market forces, fostering long-term relationships with suppliers, and intelligently aligning procurement with the larger business goals of the organization. Procurement pros are getting smarter about sourcing decisions by using AI to automate purchasing processes; create guided buying journeys; and automate intake, triage, and coordination.
How does AI actually work?
For as compelling as AI is, tech for the sake of tech accomplishes nothing in the enterprise. Procurement folks need to get strategic about how they leverage new technologies like AI to build cognition and intelligence into key aspects of their practice.
The first step is realizing that AI is just a set of tools. Learn how it works. Become aware of its limitations and liabilities as well as its capabilities, and parse the use cases where AI can actually be effective versus those that require a human. And instead of starting with a fascinating piece of technology and looking for somewhere to apply it, start with the problems you want to solve and search for solutions. And be sure it’s all aligned with your organization’s larger business goals.
So how does AI work?
AI is more or less just very fancy pattern matching. It’s based on AI models, which are algorithms that are trained on datasets to uncover patterns. These models are designed to do something you need—look for particular terms in contracts, analyze a supplier database, and so on. (This includes more advanced tasks like letting users ask for what they’re looking for using natural language.)
Data is fed into the models to “train” them. AI training simply means teaching a particular model to interpret data correctly, often by asking it to finish an incomplete pattern and then confirming or correcting the model’s guess, which improves its accuracy over millions of these guesses.
After the model is sufficiently trained, its makers can put it into production. Then, you can have the model do what it’s designed to do for your purposes—sort numbers in your list, and so on.
The more data that gets fed into the training set, the more it “learns,” and ostensibly the better the model performs.
The role of generative AI
Generative AI (like ChatGPT and DALL-E) is a unique part of the ecosystem because it’s so easy for average people to find and use, and the results are immediate. You give it a prompt, and it spits out new text or images nearly instantly.
Unlike, for example, robotic process automation (RPA), which is powerful but only for certain people and in certain ways, generative AI has sparked the imagination and caught the attention of tens of millions of people and created a frenzy race in the SaaS world.
It’s the large language models (LLMs) that have been the most prominent. But like every new technology, generative AI is in the midst of a hype cycle in which it’s often mischaracterized and misunderstood. In that sense, it’s no different than any other technology tool—or other types of AI, for that matter. You have to understand what its power and limitations are, and how it works, in order to strategically employ it in your organization.
For example, say you’re executing an RFP. An LLM could create it for you, and that saves human time and labor. But you still need a human to double check it for accuracy, and the supplier on the other end still has to fill it out, and you still have to build a relationship with that supplier in order to do business, and those are all things an LLM simply cannot do.
It’s important to understand that although LLMs like ChatGPT can feel almost magical, that’s just because it’s new technology. They’re just scouring what humans have already made and then rearranging them in a way that resembles what humans have already made. It is, however, powerful; and because it’s a tool and not magic, you can learn how to use it to enable your organization’s digital transformation.
So how should procurement professionals be thinking about using AI to push digital transformation?
It’s an AI-powered intake experience for submitting and proactively resolving internal requests. The tool has a simple interface that can be “opened” from anywhere. It’s unified, meaning anything employees submit through the Front Door, no matter where it comes from, gets funneled into a unified portal for the procurement teams on the other end. When employees want something—a document, a purchase request, client update, and so on—they can use regular language to ask for it, and the system will produce what they need from your organization’s own data and resources.
Partnering with a platform that makes specific, strategic use of AI is an important step. The AI Front Door provides an AI-powered intake tool. That’s an excellent place to start when you’re thinking of how and where to tap AI.
The importance of people
But that’s hardly where you should stop. It’s perhaps ironic that as technology like AI continues to get more powerful and can automate away many tasks previously performed by humans, including creating written content and images, the need for people has never been greater. The general term is “human in the loop.”
That is, we can’t blindly trust these AI models; people with expertise have to be able to check their accuracy and veracity, and they need to be able to step in and take over when necessary.
Nor should we look to AI to do everything for us. Rather, in procurement as in other fields, AI is a tool that should be used to complement our unique skill sets as humans, and provide us with more bandwidth to invest in those skills and focus on the correlated tasks.
There will always be things that only humans can do, like the aforementioned RFP example–particularly the relationship-building with a supplier. As Tonkean co-founder and CEO Sagi Eliyahu noted recently on the Modern Business Operations podcast, generative AI commoditizes certain aspects of work, which allows us to pay more attention to other aspects of a process when it requires a human touch.
Procurement professionals should be ambitious, however, when thinking about what to automate with AI; there’s ample opportunity.
“If it’s something that’s three-bids-and-a-buy, let a machine do it,” Tiura said. “If it’s going to be something that’s going to influence the outcome of the organization, or their bottom line, then have it be strategically sourced, and use smart tools that use AI, that use cognitive procurement to do that kind of sourcing.”
In other words, let AI do tasks that it can perform faster and better than humans, keep humans in control of output from AI tools, and leave the truly human tasks to people.
Use AI in Procurement safely while mitigating risks
Of course, it’s unwise to treat any AI, or any new technology, with careless excitement. The key with these waves of technological advancement is to adapt—and to embrace regulations as they’re put in place and when there is a need.
Furthermore, any given AI solution or cognitive procurement strategy needs to be tested; procurement teams should always understand how the tools and strategies they invest in are working, and why.
Here are some general best practices you can follow when you’re vetting an AI tool:
Ensure human-in-the-loop design, which creates transparency and enables people to see how an AI is working and step in if need be.
Use it for narrowly defined applications (even with powerful generative AI like GPT).
Understand the concerns germane to your vertical (e.g., there are different things to worry about between Procurement versus HR versus Legal).
Make sure it’s purpose-built for what you need; that is, use a tool that’s empowered by GPT, like the AI Front Door, rather than just letting everyone use GPT with no parameters or clear goals.
Education: Teach people in your organization how to use your tools properly. Some of that will be specific to your organization, and some will be part of the larger culture—for example, just as we all had to learn how to use an internet search engine, so too do we need to learn how to give LLMs prompts that will deliver the resolution we need. (That also makes it easier for your employees to follow your procedures and policies, which reduces risk.)
If necessary, gate the tool so people in your organization can’t accidentally use it destructively. (This is similar to the way we hold software licenses and permissions.)
Data oversight, such that you know where your data is going and how it would be used—which you can check by looking up the Terms of Service on the tools you use, like OpenAI—and whether or not you can opt in or out of data collection.
TLDR: AI holds a great deal of promise, but it’s just a set of tools. However, in the right hands—procurement professionals with knowledge, vision, and a strategic mandate—these technologies can help you make a huge impact on your procurement processes.
The word of 2024 in procurement so far is “process orchestration.” Everybody—from procurement experts to executives at Goldman Sachs—is talking about it. But… why? What are they so excited about? Here's why.