TRANSCRIPT:

Fabrice: All right. Today I'm joined by Dr. Elouise Epstein, Partner at Kearney, who has spent years challenging how procurement needs to evolve in a digital-first buying world. Dr. Epstein, welcome to Open Buying Conversations.

Dr. Epstein: Thank you for having me. I'm excited to be here today.

Fabrice: Thank you. Let me start right away with the first question — let's get straight to it. You've written quite extensively about digital procurement enabling business success in a disordered world. What is fundamentally broken in how indirect procurement operates today?

Dr. Epstein: Favorite question. How long do you have? That's been the thrust of my work for the last decade. Well, first — the way corporations have come to understand the cost equation has really changed over the last 30-plus years. It used to be about spend controls, and procurement was a nascent organization when I came into the business world. That has changed. It has professionalized. But in between, we had consultants and systems integrators who embedded an ideology that you could control procurement and external spend — and then tool providers came in and built tools to that specification. The disconnect is just massive. I kind of blame consultants, tech providers, and systems integrators the most, because nothing has evolved to where it needs to get to. And we are now on yet another precipice of massive change with AI. The entire construct of source-to-pay and linear procurement doesn't work anymore. It hasn't worked for over a decade, if not longer.

Fabrice: I think there's an assumption — and I'll admit I was on the systems side so I'm doubly guilty here — that buyers must come into the company system, rather than meeting employees and suppliers where they already are. Are those systems still designed for a world that no longer exists? They still persist, and I keep hearing about tweaks here and there, but are they still valid?

Dr. Epstein: Are they valid? No. Are they used? Absolutely. And that's the disconnect. My next book is going to be about this — AI is going to expose that disconnect. I'm not particularly interested in AI simply automating the enterprise; I believe that will happen in whatever form it takes. But what AI will really do is expose the inadequacies and ineffectiveness of entire bureaucracies within procurement — and other functions too — where we're running old processes on old technology and waving our arms to make it look like we're meeting objectives. We're not set up for success. At some point the floor just drops out beneath all of this. That's why I'm pushing so aggressively — I'm genuinely concerned for the profession if we don't get ahead of this.

Fabrice: What's the biggest bottleneck? Is it fear of change? Lack of imagination? Too much lobbying from the wrong consultants and system integrators? What's the biggest hurdle?

Dr. Epstein: I think it's fear. There's that old adage: nobody gets fired for buying IBM. Well, there's a procurement corollary: nobody gets fired for choosing SAP. Part of it is fear, part of it is inertia. You have all these S/4HANA migrations happening right now, and it just seems like the safe thing to do — but it's counterintuitive, because it's not the safe thing to do. We all know the big ERP platforms and the big source-to-pay platforms aren't going to get it done. I talk to clients constantly who are being pushed to adopt Ariba — old Ariba, not even the new version, which won't be much better — just because they're "an SAP shop." If you're a CPO, you have to step way outside your comfort zone to raise your hand and say, "I want something different." And there's plenty of uncertainty about what that something different actually is, which makes it even harder to stake your career on.

Fabrice: Should we expect the larger vendors to lead the pack on AI — at least giving us some transformation, even if it's still within the SAP world — or do you see a real role for emerging players to shake this up? How do we make the change happen?

Dr. Epstein: I don't think it's going to be the legacy players leading this — not because of us, but because their role in the enterprise is already getting diminished. The CIOs who genuinely get on board with AI and drive it into the business will remain relevant, but that's a small minority. CIOs are largely not ready for this. The entire Chief Digital Officer role exists precisely because CIOs haven't evolved — if they had, we wouldn't need CDOs, because we'd have actually done digital and moved into AI. And at the best-case scenario, AI completely eliminates 90% of IT resources. More fundamentally: at a certain point, you are simply not going to be an AI-native organization running SAP. Full stop. That doesn't mean you have to rip it out — just be clear about what it is and what it isn't. Think about drug discovery or producing biologics — the barrier to entry for you and me to compete in that space is getting lower by the day. If you're a large pharma company just lumbering along with "SAP AI," you're putting yourself at massive risk — and not just the enterprise, but everyone who works there.

Fabrice: What's the risk, concretely, for a large organization that keeps rolling the dice and just layers AI on top of an existing platform?

Dr. Epstein: You're going to become increasingly inefficient while everyone around you becomes more efficient. OpenAI's CFO Sarah Friar talks about how her team is 18% the size of comparable companies because they're building everything with AI. You'd expect that from OpenAI, of course — but if that becomes a broader model, and you're not thinking that way, you're heading toward massive inefficiency. And I'll be the first to admit I was slow to get on board with AI — I was highly skeptical. But what's become clear to me is that AI doesn't care about your feelings. It's going to come in and eviscerate processes — not apocalyptically, just efficiently. If you have 10 people managing a process, AI is going to eliminate the need for those 10 people. And the real risk isn't just to the enterprise — it's to individuals. If you're working somewhere that's pushing SAP and calling it AI transformation, and you eventually leave or get let go, you're going to be competing against people who've been immersed in OpenAI, Gemini, or Anthropic. Walking into an interview saying you're a platinum expert in SAP AI while the person before you is a wizard in Gemini — and Gemini could easily eat SAP's lunch — is not a position you want to be in.

Fabrice: So the process is broken, AI is changing this fundamentally. If we narrow it down to indirect procurement — and you know this space better than anyone — if you were advising me and we had to rebuild an entire indirect procurement suite from scratch, how would you design that with AI at your disposal?

Dr. Epstein: I think it's actually quite simple, and I have an architecture slide I can share with the show notes. I think of it in layers. The first layer is the ERP — that's a transactional system of record, and I have no issue with it as such. Just don't try to put all the business logic in the transactional layer. Keep it as a transactional layer. On top of that, a data lake — we all agree on that: land all your data and make it accessible. Then an AI platform — Gemini, Anthropic, OpenAI, and others that will emerge. Then, crucially, a procurement operating system — one thing you go to market for, one thing you buy for procurement, that provides the framework and container, leverages the AI platform, lets you build agents and custom solutions, and comes with sourcing and the core basics built in. Between those four elements — ERP, data lake, AI platform, and procurement OS — you can solve virtually any problem: direct, indirect, risk, you name it. That's how I see it.

Fabrice: And for the procurement team and their boss — typically the CFO — is the tension still going to be "CPO, go negotiate as much savings as possible," or is it going to shift more toward process efficiency? What's the new economic footprint?

Dr. Epstein: I'd flip it around and talk about what I call "blackboxification." Imagine the procurement operating system I just described — ultimately it becomes a black box. You'll have five or ten people managing that black box, and they can give the system instructions: get 10% savings, identify opportunities — and the black box just keeps doing its thing. The people previously doing that transactional work now move into the business to do real risk analysis and supply market intelligence. So the CFO's push for savings is absolutely still there — the black box actually empowers that. But from a career perspective, those of us in procurement will have the ability to step over into the business and make a genuine strategic impact.

Fabrice: That's a brand new perspective. I suppose this new framework will replace the spider map. When can we expect to see it?

Dr. Epstein: I'm happy to report the spider map is retired — enjoying retirement and having cocktails on the beach. My focus now is much more on people and culture. I think two things are going to happen. A handful of companies will build out the procurement OS — and they'll be net new players. I wouldn't crown anyone yet, and I'm not in the business of crowning companies anymore. But at some point we'll have a couple of legitimate, AI-native players. It won't be six, it won't be ten. And none of the legacy source-to-pay vendors will make that leap — let's stop pretending they will, because they won't. The other element is a whole set of productivity tools and productivity agents that are net new — I'd put Blue Bean in that bucket, along with a dozen others, and I'd expect another couple of dozen to enter the market over the next few years. That's exciting. The market is going to look very different from what we've known — not source-to-pay as the center of gravity, but a procurement OS surrounded by a constellation of productivity tools and agents.

Fabrice: Let's close this conversation with some advice for practitioners who are about to face these changes. Not providers — practitioners. What would you advise them to do to embrace this shift, transform their organizations, transform their careers, and create that leap in productivity?

Dr. Epstein: I'm going to draw a line. For those of us over 40: stop and learn AI — go all in on it. It's hard and it's going to take us ten times longer, but you have to start now. Train your brain so you can become a leader in this AI world. You have no choice, unless you're within three years of retirement, in which case you'll be fine. For those under 40: I'd expect you to have built 20 agents — and be able to tell me how those agents are improving your personal productivity, whether at work or in your personal life. I'd want to know how adept you are across different AI models, because those are the capabilities that will matter going forward. And I'd want to know how strong your critical thinking and relationship skills are, because those are the human skills AI won't replace. If you're at a company that thinks SAP is the pinnacle of digital innovation, you are stunting your growth. AI is going to expose that SAP, for all the hype, has been largely a massive cash grab for big tech, big systems integrators, and consulting firms. You have to take control of your career. And just one more thought: for the last 30 years, if you were an Excel wizard you were considered untouchable and brilliant — and you probably made a lot of money. The corollary now is that if you're an AI wizard, you will at minimum be competitive, and in the short term likely be very successful.

Fabrice: Dr. Epstein, thank you very much for sharing your views on how this world is going to change. As a 50-year-old myself, I'm looking forward to continuing to learn. Thank you for your time.

Dr. Epstein: Thank you for having me.