Opportunities and limits of artificial intelligence in procurement

erstellt am: 12.06.2023 | von: Friederike Kastl

ExecutiveTalk with Dr. Marcell Vollmer

He was CPO and Chief Digital Officer at SAP, partner at the Boston Consulting Group, and today leads a group of businesses in the health care sector and he is considered a futurist and business influencer for digitization, especially in supply chain management. We spoke with Dr. Marcell Vollmer about the opportunities and limitations he sees in artificial intelligence in procurement.

Where do you see artificial intelligence in SCM and procurement today – in which areas are AI technologies already being used?

Marcell Vollmer: In supply chain, there are already a lot of applications based on machine learning: Demand forecasting already works quite well, as does inventory planning. In logistics, algorithmic systems can be used to optimize routes. And in procurement and sourcing, I find it particularly exciting that large language models such as ChatGPT can be used in the future to better evaluate the quality of the many catalogs with very different information and to prepare the best product offering for the end user. Likewise, there will be call-off contracts with intelligent forecasting. As in all other areas, the world is on the move here.


What limitations do you see?

Marcell Vollmer: We’re all talking about artificial intelligence right now, but in the end, we’re talking about algorithms – albeit at a whole new level of quality with ChatGPT and other AI solutions. However, intelligence is not yet truly ‘human’. Let me give you an example: If an important supplier fails, then very relevant decisions often must be made in a very short time. You need a substitute for product, service and supplier. But as it stands today, AI can only suggest options based on rule-based parameters. Truly determining who exactly fits you now – it can’t do that. People must decide that. Which is not to say that this may not be significantly different in a few years. ChatGPT hints at the potential in this area as well.


How will CPOs manage in the coming months not to fall for ‘hype’ or false promises – but to keep a cool head in the new AI jungle?

Marcell Vollmer: It’s actually the same as in the digitization wave: follow a clear concept. And that hasn’t changed at the core: Start with a vision. Think from the perspective of your stakeholders. Have a clear business case. Set up a process for this. Then look for a technical solution. Focus on execution. Experiment outside the box, but not by the hundreds. If you have a hundred priorities, you will end up with none. And finally: Always remember – no matter how good the technical solution is – it’s useless if you can’t bring people with you. Unfortunately, I have had to experience this myself time and again in my roles as an operational manager – and I have learned from it.


If you had one wish: Which problem in procurement, would you like to see solved with AI – and which very personal challenge?

Marcell Vollmer: From my point of view, one of the biggest problems for enterprises is data quality. Catalogs are only one example. But it also goes even more mundane. There’s the old joke between procurement professionals: Look at your own data to see how much spend you have with a supplier. And then ask the same question to your supplier. It’s almost impossible to come up with the same number. If AI could help create more transparency here: that would be great. And personally, as a German and European, I would like to see not only our business leaders but also political decision-makers approach these new technologies with openness. And not put regulation before opportunities. However, this problem can only be solved by human dialog – not by AI.


Meet Marcell Vollmer in person at Crowdfox’s masterclass “How AI is fundamentally changing procurement” at the Procurement Summit. For more information about Dr. Marcell Vollmer, please follow this link.