Transformation as Strategy: From Data to Dollars: Unlocking AI’s True Value in Procurement

00:00:06: Hello everyone and welcome to Transformation as Strategy, a podcast brought to you by the experts at Roland Burger Americas where we explore fresh perspectives and practical strategies to help business owners create a lasting competitive edge.

00:00:20: I'm your host Mackenzie Pudisey And the topic for this episode is The Current State of AI in Procurement.

00:00:27: Joining me today's conversation is Max Gustafsson A Principal At Roland Burger.

00:00:35: Thank you, Mackenzie and excited to be here.

00:00:37: Yeah Max do you mind letting us know a little bit more about your role at Roland Burger?

00:00:41: And your day-to-day in the industry?

00:00:43: Sure yeah so as mentioned I'm a principal In our operations practice Located just outside Washington DC.

00:00:51: But yes part of Operations i focus on really all manner Of cost transformation topics And one of The most integral functions that's Part of That is procurement.

00:01:04: So we'll be talking today through the lens of how AI is changing, how you do procurement.

00:01:13: Excellent thank-you!

00:01:15: Okay let's dive into this.

00:01:16: I know procurement as often described it one a few business functions where AI can drive both operational efficiency and direct cost savings.

00:01:25: so what makes procurement uniquely positioned to benefit from AI compared other enterprise function?

00:01:32: So I think procurement is distinct in that, you know it sits at this intersection of data decisions and money dollars.

00:01:43: so unlike most functions.

00:01:44: It manages large volumes of structured spend data but also unstructured In highly repeatable processes kind of creating strong ideal conditions for AI to automate tasks And improve efficiency.

00:02:00: At the same time, procurement does directly influence decisions that is from procurement.

00:02:07: Influence supplier selection pricing and demand aggregation enabling AI to generate tangible cost savings.

00:02:15: so I think it's this dual link between process efficiency which really any function can benefit from And bottom line impact That i do thing.

00:02:26: as specific to procurement one of the most compelling areas for AI value creation.

00:02:33: Max, I think you've really set the stage there and it sounds like... There's a lot going on in the industry!

00:02:39: You're living it, your kind of observing it.

00:02:42: so with that perspective i wanted to ask how are todays procurement leaders approaching AI?

00:02:48: Are they viewing primarily as productivity tool?

00:02:51: is it strategic decision making asset or something else entirely?

00:02:58: Yeah, so I would say first today procurement leaders are still in the early stages of AI adoption.

00:03:04: You see many focusing on testing use cases building proofs of concept and running pilot programs rather than any kind of big large scale deployments.

00:03:15: So to your question AI Today is primarily viewed as a productivity enabler enabling helping teams do what they already Do but more efficiently by automating manual in repetitive tasks.

00:03:27: So the immediate benefit that leaders see is improved productivity through having, you know capacity that has now freed up within their team being able to accelerate the time at which it takes to do activities and then allowing your teams to focus on higher value activities rather than these repetitive tasks so that really kind of lays a foundation to expand into more advanced values generating use cases over time.

00:03:57: In these more advanced use cases, for example what are they?

00:04:00: They can enable better decision making by improving how procurement analyzes and acts on that data such as identifying savings opportunities within specific categories of spend.

00:04:16: Optimizing your supplier selection in what suppliers you're even bidding to What regions are you looking at?

00:04:24: And then strengthening how you negotiate with the suppliers.

00:04:27: So it can provide just deeper spend visibility and real-time risk insights, but also allows procurement to make more proactive in informed decisions.

00:04:37: And then The last thing I want to mention is that It broadens the scope of what procurement teams can address.

00:04:43: so Again your team today has a fixed capacity That maybe you really can't go and bid every category as frequently as you want too.

00:04:53: With AI It frees up some of that, again the bandwidth to focus on higher value activities allowing better management things you weren't able to monitor so closely in past.

00:05:05: Things like your tailspin or small vendor spend complex categories like MRO spare parts and other related supplies like packaging.

00:05:17: So as a result I think AI... The first thing it does is shifts procurement from simply Executing processes faster and more efficiently, but then it kind of moves you to More actively driving better outcomes.

00:05:31: And unlocking new sources of value.

00:05:34: Excellent I think It sounds like you You've given us a sense of sort of what's being done now?

00:05:41: Uh, and i know A lot Of people are starting To implement Now in these last few years some AI initiatives.

00:05:48: There is some upfront costs there.

00:05:50: There's some learning curve, absolutely And so I believe that it's an in an infrastructure investment.

00:05:58: from what i'm hearing From you are there any pitfalls?

00:06:01: That limit ai impact in procurement.

00:06:04: and You know why do these challenges persist as people are starting to get in and roll out this investment?

00:06:11: Yeah, I mean I think There's several pitfalls.

00:06:15: So I don't think what all mentioned is not exhaustive, but ii-I Think they're kind of common stumbling blocks that uh the companies run into.

00:06:24: and i'd say The first Is around the data?

00:06:27: Um so really this starting point Of what data are you working with?

00:06:32: um And that data You know does need to be.

00:06:35: we call kind of AI ready To be able to do something With it.

00:06:40: so That's one of the first areas.

00:06:43: I think company stumbles.

00:06:44: they say well our data is too disjointed, inconsistent it's not all connected so we don't have one place that can start to pull from and analyze.

00:06:58: So really I think thats the first area of what we like to preach about.

00:07:03: you need a robust semantic layer to connect and standardize your data assets.

00:07:09: A semantic layer basically has standardized framework which translates raw, complex data into business-friendly terms.

00:07:17: And it usually sits between your data storage, data warehouse, data lakes and where you're consumption tools lie.

00:07:25: so its kind of a unique universal translator or hub and spoke that interfaces with the rest enterprise.

00:07:34: So Data is number one.

00:07:35: I think thats kinda stumbling area.

00:07:38: The second is thinking that narrow thinking AI is solely gen AI.

00:07:45: Gen AI is very popular, you know?

00:07:48: You hear it a lot and it is impressive.

00:07:50: the things that you see...you know..the image-based capabilities that has but again....that's just one piece.

00:07:58: I think that kind of limits the scope use cases and misses broader analytics in optimization opportunities.

00:08:06: so i do gathering some of that data from a broader range of sources, such as from drawings and CAD files.

00:08:17: From PO's invoices through more modern AI OCR techniques to interpret and translate the visual data again input into that semantic layer basis I talked about before.

00:08:32: so i think it can help feed the data issue.

00:08:35: but its not only thing you could do with AI.

00:08:39: And then the third last point I'd mention is that many organizations are very focused on cause and effect.

00:08:46: I invest, I get quick wins.

00:08:49: so there's kind of an over indexing on quick win results at the expense of building a longer term plan that can lead to transformative value.

00:09:00: So i mean That gets compiled in by An absence of really AI ready operating system.

00:09:07: Sorry, operating model where roles processes and governance are not adapted to fully embed AI into procurement workflows.

00:09:16: So you do also need kind of a good hot model that's ready too.

00:09:21: basically accept longer term transformative plan.

00:09:27: so it is really either or situation.

00:09:29: but I think he needs both quick wins because you need those to build the confidence in the organization that there is value worth pursuing, but also a longer term roadmap to unlock.

00:09:42: The true transformative value that I can offer

00:09:47: excellent.

00:09:47: it sounds like people do needs sort of get their data and order get their house an order have a good perspective on what's realistic short and long-term.

00:09:57: And so actually, I want to shift the conversation from what's happening now in these last few years as AI has become available broadly and where we're going.

00:10:10: So as this AI capabilities continue to evolve.

00:10:13: which procurement processes are generating the most compelling results today?

00:10:19: Where do you see the biggest opportunities emerging over the next few years Max?

00:10:24: Yeah!

00:10:28: The issue I mentioned earlier with having a strong basis of data is kind of the first place.

00:10:35: I've seen AI be deployed to go and address, try and piece this together.

00:10:40: so you do see pretty compelling results from spend categorization in analytics using AI either suggest taxonomy to categorize your data or taking a taxonomy that the company uses and then just feeding in vast amounts of data, asking it to categorize your suggested taxonomy.

00:11:08: So I think that's where you're seeing work well rapidly pretty effectively today but looking forward once you have that in place... The bigger opportunity emerges is organizations build off of that dead if and out foundation in workflows, bigger things.

00:11:29: So you can move beyond analytics to more autonomous execution really.

00:11:35: so this is a thing like deploying agents.

00:11:39: they can actually directly interact with suppliers.

00:11:43: They could interact.

00:11:45: just do simple things like manage contracts make them kind of monitor the performance of a supplier and obviously communicate with the supplier to react to that.

00:11:58: But again, even we've seen agents concepts look at supporting negotiations across large swaths of suppliers so really doing negotiations at scale which is something you could never do with just humans trying to negotiate.

00:12:19: So you get more into kind of the agentic solutions that would evolve, beyond just the visibility and analytics insight that we get today—the capability does drive broader execution in value creation in procurement over time.

00:12:37: That's exciting.

00:12:38: I think that agentic layer can be really powerful, especially in procurement and supply chain.

00:12:44: when there are so many moving parts people in different regions quantities prices like you said negotiating that is going to be an interesting evolution.

00:12:55: now max of people out their end.

00:12:57: they're just beginning this AI journey in their companies and the corporations.

00:13:02: what should a procurement leader prioritize first to ensure they're building a strong foundation rather than just chasing the latest tech trend?

00:13:12: Yeah.

00:13:13: And again, this where-to start depends on their company and where the starting point is for the company.

00:13:20: So really what we like to do is first understand where are companies at today through maturity check so that would assess the readiness of AI adoption by getting a pulse of the organizational capabilities, as well as the technical.

00:13:38: I mean people focus a lot on what platform are you using?

00:13:43: Does everybody have licenses access to this?

00:13:46: but it's really...I think that PEOPLE element is equally important because they're the ones we have use and interpret results.

00:13:54: so first is getting a maturity check.

00:13:56: we call them The second.

00:13:58: We recommend developing and testing targeted AI use cases, focusing on high impact opportunities that can demonstrate tangible value.

00:14:10: So this can be something as foundational is data categorization and harmonization like I mentioned before but it could expand to more of an agentic solution.

00:14:20: if you actually already have a good data foundation in your you're ready for.

00:14:29: Once you've tested some use cases, organizations should invest in some sort of enablement phase.

00:14:37: So usually through the use of a pilot or pilots to stand up that tech stack build the internal capabilities within the organization but whilst kind of scaling into more and more successful use cases.

00:14:52: so your going beyond just use case now really piloting it specific area.

00:14:58: And then finally, I think if you really are trying to get that full AI first scenario we recommend defining a clear AI acceleration strategy.

00:15:12: You know start including a prioritized roadmap governance model.

00:15:18: what's the impact?

00:15:19: We expect to get from this to ensure sustained value creation.

00:15:22: so This sequence approach that organizations are building the right data, the right capabilities and operating model to scale AI effectively rather than just pursuing fragmented ad hoc initiatives.

00:15:43: I think that's important.

00:15:44: The step by step is... You often want to jump a step because AI is moving so fast, it sounds exciting.

00:15:53: The promises are real they're high.

00:15:55: So I do think that's great advice.

00:15:57: Max thanks for that.

00:15:58: and you mentioned before some kind of like short-term wins long term strategies?

00:16:03: I wanna ask you little bit more about that.

00:16:06: with AI moving so quickly What should leaders be doing to achieve some of those short-term AI wins and what are those?

00:16:14: You know, if you can give us some examples with sort like...what are some wins on the table that maybe have seen people could go for.

00:16:21: And then how do they lead into a longer term organizational value?

00:16:26: Yeah so again I think it's not an either or situation.

00:16:32: The balance between capturing early wins and building towards transformation.

00:16:39: these are not competing priorities.

00:16:40: So in the near term, I think pick something that is obviously achievable within kind of the realm you can control as a leader.

00:16:54: so whether it's just your function or if its within series functions maybe procurement and product development.

00:17:03: but again The good place to start is automation, analytics data categorization because it's critical to kind of show and demonstrate that value early on.

00:17:15: It's something that's tangible That people can understand.

00:17:21: But yeah by doing that you would build momentum You create organizational buy-in And I think many companies are already seeing strong results from piloted programs in targeted use cases.

00:17:34: I'll just buy it.

00:17:36: They can see that they can execute faster when they have this, and they can measure very empirically what the performance gains are.

00:17:46: so again kind of picking things to allow your team do things faster or a broader scale or being able to address more than you could previously is good demonstrations too with hey!

00:17:59: This has value.

00:18:03: leading organizations don't just stop there.

00:18:06: They're pulling ahead, or at least those that you see kind of pulling ahead are treating.

00:18:11: AI is not just a tool for efficiency but as it's a catalyst for broader business transformation.

00:18:18: so they use these early wins the foundation to drive deeper changes particularly in data quality process standardization.

00:18:27: and then you get into cross functional integration which probably, yeah one of the bigger unlocks but it's just is much harder to go straight too because you need each function.

00:18:40: To be kind of familiar and have some um yes some proficiency in already functioning in an AI environment.

00:18:50: so I think it's important to kinda just build up the comfort.

00:18:56: Cross-functional integration is where you can start to get really big collaboration boosts Again deploying these tools we're talking about.

00:19:05: so Yeah, I think Practically just it.

00:19:09: I guess they recap at all.

00:19:11: You know use quick wins to prove value and build momentum.

00:19:14: Invest in data integrity and harmonization to enable scale Align the leadership an operating model around kind of a singular AI vision so that everybody's aligned on where you're going to, and then start to embed AI across the workflows rather than just isolating it in pilots.

00:19:35: So I think those organizations that succeed are the ones that avoid this kind of pilot purgatory using short-term results to fund and justify longer term transformation which we really need AI into how procurement operates.

00:19:55: Max, I might have to borrow that term from you the pilot purgatory?

00:19:58: I feel like yeah definitely want to avoid.

00:20:02: don't wanna be trapped there

00:20:04: exactly.

00:20:05: well You've got us a lot of great advice so far.

00:20:08: i think he's been really concise and clear about How we can start where We can have some Of those wins short or long-term in what were going.

00:20:16: And so I want to ask you one final question here, a question we'd like to ask all of our experts on the show.

00:20:22: What is One Piece Of Advice You Would Give To Businesses That Are Looking To Stay Ahead Of The Curve and Compete Rather Based On The Trends That We're Seeing In Procurement in twenty-twenty six?

00:20:36: Yeah So i think that first thing Ill say Is Not To Get Discouraged.

00:20:44: AI is everywhere, it's going so fast.

00:20:46: We must be behind and really that's a misnomer.

00:20:49: I think most organizations are still early in their AI journey.

00:20:54: In fact what we've seen as more than half or still testing and piloting phase Or they're just starting to build the road map.

00:21:02: So you're not behind?

00:21:03: The key is move with intent rather then speed.

00:21:07: Put clear plan into place start with high impact use cases, build the right data foundation and then scale from there.

00:21:15: So just it's really the basics that you want to establish in demonstrate first.

00:21:22: And yeah I think organizations at Wynn aren't the ones who are chasing latest tools but they take a structured deliberate approach embedding how AI can work within their procurement processes.

00:21:38: So how they really operate using AI?

00:21:43: In short, you're not late but now is the time to start with a clear plan and build momentum.

00:21:51: Beautiful Max.

00:21:52: thank you so much for your valuable insight throughout this conversation that wraps up our discussion today.

00:21:59: I truly appreciate those strategies.

00:22:01: insights help business owners navigate trends in the procurement landscape.

00:22:06: So again, thank you so much for your time.

00:22:08: Thanks for having me

00:22:10: absolutely and For our listeners out there.

00:22:12: if You'd like to learn more about what max in the team at rollenberger americas are working on that?