AI in procurement – especially generative AI – is going through the hype phase at the moment. But make no mistake. In one form another, it’s here to stay.
It’s reasonable to expect that its wings will be clipped by some data privacy and confidentiality concerns, especially in large enterprises. I suspect there will be a lot of challenges from information security (IS) and legal departments.
However, it’s proven ability to cut down the time and resources needed to complete many everyday tasks is not to be underestimated. If you’re ignoring it, or saying it’s garbage, then do so at your peril. Even though some of your concerns may be reasonable, do not bury your head in the sand.
Could AI in procurement make us irrelevant?
AI is not going to just go away. It’s not a latest fad, even though some aspects of it may be overhyped. You’re going to lose ground pretty soon against your competitors if you don’t understand both its benefits and its limitations.
So, let’s look at where it will help us, where it could be a threat, and whether indeed AI in procurement will make some job functions and processes irrelevant.
What is Generative AI and why is it different?
First, it’s worth quickly touching on the difference between Generative AI and the already established AI which is / was already present in procurement technology applications.
A recent Forbes article manages to do this succinctly and in plain English:
The main difference between traditional AI and generative AI lies in their capabilities and application. Traditional AI systems are primarily used to analyze data and make predictions, while generative AI goes a step further by creating new data similar to its training data.
Generative AI is most commonly associated with apps such as ChatGPT. These use Large Language Models, or LLMs. These predict the output based on the data they have been trained on, and suggest the next word or sentence based on what they have been prompted. LLMs are just one form of generative AI though.
As the technology develops and matures, many more use cases will gain traction and become popular within everyday use. Gartner has also put together an interesting, well written article on the topic which I would highly recommend as an introductory guide.
Now, I’m not a mathematician or a computer scientist, and I’m not going to attempt to describe what else is out there. It’s way beyond my pay grade. Maybe I need to get Melissa Drew back on the podcast again, given how much has changed since we last spoke about AI.
Suffice to say, Generative AI is a quantum leap vs. “traditional” AI.
How is machine learning different from Generative AI?
Machine learning is a subset of AI. Its algorithms can handle a wide range of tasks to aid different areas of business. It gets more “intelligent” based on the increasing amount of data sets you feed it.
The most recognisable example of machine learning in everyday life is the algorithm used by services such as YouTube and Netflix.
What about in the procurement technology space?
Autonomous sourcing and autonomous negotiation applications, as well as spend analytics tools, have all used machine learning algorithms to train their bots on the data.
The case for Generative AI in procurement
Nowadays, with the opportunities that AI in procurement offers up, there’s no longer a need to grind out research or analysis manually.
Nor do you need to draft pretty standard documents. Generative AI can put together a perfectly functional NDA, or standard RFI / RFP document. It will also be able to assist in monitoring supply chain disruptions by being able to quickly summarise foreign language news sources and weather reports.
Younger team members, who are using tools like ChatGPT at home, will expect this level of efficiency in the workplace. Pretty soon, you’re going to struggle to recruit the best people into roles if you’re not taking advantage of Generative AI.
ChatGPT for example costs $20 a month for a premium license. It can perform, or drastically reduce, the workload for a lot of tasks that are not only administrative in nature, but also those being performed by Buyers and Category Managers.
Now, I’ve played around with these Gen AI tools. I use them in my business, but I’m by no means an expert. AI can now write my podcast show notes for me.
Previously, I had to pay someone to listen to my episode and then summarise it into some notes. Just think that even in my small business, it saved me $30 in production costs per podcast episode. And this is just one example. It does a lot more for me in the background too.
So, how big are the opportunities in a large enterprise, with all of their well-known inefficiencies and banal bureaucratic requirements? Applying that same curious and can-do mindset can also drive efficiency gains in the daily grind of the average front line strategic procurement role. AI will be the darling of the Category Manager or Strategic Sourcing Manager.
What procurement roles will AI replace, and where are its limitations?
Going back to the original question, the application of AI in procurement will replace a lot of transactional and administrative work. So, if we consider whether AI will make procurement roles irrelevant, then I would summarise it this way.
It’s not going to make strategic procurement irrelevant, but it is going to make a lot of transactional, administrative, research and analysis work pretty irrelevant.
Consider how guided buying already allows stakeholders to bypass dealing with a purchasing clerk when they want to buy something simple. The use of catalogues or e-procurement software was the first step.
Now, you’ve also got automated sourcing and supplier discovery tools in addition, that can cover a lot of non-repeatable, one-time spend too. And yes, that does include the negotiation piece if you look at something like Pactum. Then you’ve got digital procurement tools like Fairmarkit on tactical sourcing too. And, last but not least, you’ve got AI-driven supplier discovery tools on top such as Matchory, Alpas and Scoutbee.
So, forward thinking companies have already eliminated the need for someone to call tail end suppliers to horse trade with them over a 2 – 3% discount on a on a one-time purchase. Or, indeed, to spend hours googling looking for the right supplier for a bespoke purchase request.
Let’s now imagine this a couple of years down the line. It doesn’t really take much imagination to visualise that tactical and operational activity will be able to be done by a machine or by a robot.
If you’re in a tactical or administrative procurement role, AI is going to make your job irrelevant.
That’s the bad news.
But….maybe it’s good news. If you’re stuck in the weeds and want to add more value in your job, there is a golden opportunity to up-skill.
Why I think AI will help Strategic Procurement Pros to shine
So, let’s look at why I’m bullish that AI will not going make procurement irrelevant. At the more strategic end of procurement’s roles and responsibilities, it will give us massive opportunities.
Procurement is always going to be relevant as a problem solving, value adding activity to the wider business. That’s not just about delivering cost savings. It’s about how we can add value to the top line as well as the bottom line.
For the foreseeable future, AI is not going to replace a lot of the soft skills and more people-focused aspects of the procurement profession. It can’t market our value to the wider business in the way that a very stakeholder centric and communications-focused procurement organisation team can do. Likewise, it can’t build and nurture key supplier relationships.
AI can pull data together from a number of disparate sources and look at the potential risks and disruptions that could be out there in the supply chain. That’s great. It saves us from doing a lot of that work.
But it can’t build a story around that data. It doesn’t calculate the knock-on effects to the wider business in relation to the dangers that are lurking out there. A great example of this is using AI-driven guided category strategy software. AI can do a lot of the heavy lifting, but it can’t present an articulate story to the business.
For the foreseeable future, you’re going to need a human with communication skills to present the findings to the business. And you’ll likely need a data analyst to be able to tell the AI what it needs to look for or measure. In that sense, AI in procurement is really going to be an enabler, assisting human capital to focus their activity on those tasks which drive value.
Communication, supplier management, execution, and implementation of strategies are not tasks which AI can do. It can, however, be a real enabler to help us to get very specific data sets in shorter periods of time that can funnel those discussions and those decisions.
What to do if your procurement role is automated
So in summary, to what extent will AI make procurement irrelevant?
Well, P2P and tactical sourcing is pretty soon going to be automated or self-serve. It already is in a lot of forward-thinking organisations. But as time goes on, we’re going to be moving alone that bell curve beyond just the early adopters.
A lot of back office functions will go. If you’re doing a fairly tactical or administrative procurement role, you need to think, OK, how am I going to make myself future-proof for this revolution that’s coming? How are you going to develop yourself and your skillset to make yourself relevant?
That doesn’t necessarily mean aspiring to be a Category Manager. I personally think we’ll see fewer Cat Mans in future procurement organisations too. That’s a whole topic for another article.
Other roles that emerge could be a good fit to your skill set. If you’re very good at communicating with suppliers and with internal stakeholders, a procurement business partner role could be one path. Especially if you don’t enjoy negotiation and running sourcing events.
If you’re good at numbers, then maybe you could grow into being a data analyst. Or you could focus on the sustainability and responsible sourcing side of procurement, which will become an increasingly more prominent part of the work we do as procurement professionals.
Will AI in Procurement kill off some roles?
So yes, procurement teams are probably going to be smaller. Certain functions will become irrelevant thanks to AI, but it won’t eliminate procurement as a function.
On balance, procurement teams will probably be a little bit smaller in terms of total number of FTEs.
The last point I wanted to make here may be a little bit controversial and counterintuitive.
I also see a comeback of administrative roles too. Especially for admin specialists who know and understand how to use prompts for Gen AI applications such as ChatGPT.
Let’s say you’re in a tactical procurement role now. Perhaps you don’t feel you can up-skill to do something more strategic, but still want to stay in procurement. Then, perhaps look at learning how these prompts work for generative AI applications.
Will AI make procurement irrelevant? No. However, we must keep an open mind and embrace the technology, rather than fight it. However, I see more tactical, transactional, and administrative side of procurement being automated.
I will finish on a very positive note. I think this is one of the best things that could ever happen to procurement.
Why? Because it’s going to get us out of the weeds. It’s going to get us into a lot more added value and data-driven conversations with stakeholders and suppliers.
We’ll be able to put overly emotional, subjective stakeholders (or suppliers for that matter) in their place. Thanks to AI, we’ll easily have the time and resources to access the data to prove them wrong.
What’s more, AI will make the end-to-end process of pulling, analysing and presenting our case much faster. The future is bright.