Process mining: Industry leaders explore its past, present and bright future

In the past 20 years, process mining has evolved from an academic discipline to essential business technology and is poised to become the trusted, daily digital assistant for every business.

In March 2023, Gartner® published its first Magic Quadrant™ for Process Mining Tools, and not long after, Alexander Rinke, Celonis Co-Founder & Co-CEO, and Rudy Kuhn, Lead Transformation Evangelist at Celonis, sat down to discuss the evolution of process mining, explore recent advances in the technology and share their insights on its potential for the future. The pair also talked about how Celonis is pushing the boundaries of process mining with innovations like object-centric process mining, a consumer-grade user experience and an open platform strategy.

“Process mining has come a very, very long way, but when I think about where our architecture is going, where this whole space is going, I think there's so much potential,” Rinke said.

“The moment I saw process mining I realized, if this really works, it's going to change the way we, or at least I'm looking at business processes forever, just the possibility to take digital footprints from almost any kind of IT systems, push the magic button and then really see the process, that was stunning,” Kuhn added.

The following is a transcript of the conversation edited for readability.

What are your memories of the early days of process mining?

Rudy Kuhn: The early memories of process mining. Oh my, it's really a long, long time ago. But the moment I saw process mining, I realized if this really works, it's going to change the way we, or at least I am looking at business processes forever. Just the possibility to take digital footprints from almost any kind of IT systems, push the magic button, and then really see the process. That was stunning.

Alexander Rinke: It's funny, I still remember when I saw it for the first time. I actually didn't know anything about businesses. I hadn't worked as a consultant or any of that, so I actually thought that this would be a great way to generate simulation models for businesses from a mathematical standpoint. As a student and academic I thought, "Well, people would probably understand their processes, but now they can simulate them." I quickly realized, "Okay."

Rudy Kuhn: You couldn't believe that people actually don't know what they're doing.

Alexander Rinke: Exactly, don't know what their processes are.

Rudy Kuhn: That's right.

Alexander Rinke: And then I remember really getting fascinated and traveling through Germany for the first years like a madman, and I think the first two or three years I probably averaged far over 100,000 kilometers by car every year going to customers every day telling them what process mining is. And I think that's where we met too.

Rudy Kuhn: Yeah, right, from my previous job. Of course I was related or I knew a lot of CIOs and people from the IT business, and I still remember the look on their face when I told them, "Hey, we only need data and we can show you the process." "No, that doesn't work. That's not possible." "I will show you." Yeah, I don't know if you remember the very first process mining product we had, that was before we actually met, was a very, very early product from the Netherlands, the country where process mining was invented. And they had almost no statistics.

Alexander Rinke: Yeah, exactly.

Rudy Kuhn: I think that was the one thing Celonis did right from the start, that you guys were capable of combining the process visualization with the statistics. You see something weird in your process, you click on it, you see the statistics or you'll see something weird in the statistics, you click there, and you see the root cause. You'll see the process that actually created these weird numbers. And that was probably really the change maker or the game changer, that for the first time we could show people, "Hey, look at this. You have some compliance issues over here." And immediately the next question was, "What's the value? How many invoices? Who are the suppliers? Show me the numbers." And probably this is what initiated the success of Celonis, the capability to visualize the process and show the statistics.

Alexander Rinke: I agree, that was a big thing. Because you needed to really make it valuable for businesses, and that was part of that.

Rudy Kuhn: Yeah.

Alexander Rinke: When you think back to those days, how do you feel like the reaction's different today when you talk to people? Is it still the same?

Rudy Kuhn: So we have a different level of maturity around the world, and of course big companies using the major ERP systems, they know that their processes can be improved and they are constantly looking for new ways to understand what the challenges are and really fixing the problem. So they are familiar with process mining. Smaller companies are not, but I truly believe that process mining, for some it's already a commodity. So it's not a question, "Should we do it or shouldn't we do it?" It's just to what extent. And for smaller companies, I think this is going to be more common and more accepted in the future.

Alexander Rinke: Especially as we develop more and more intelligent systems, AI, and I think about, we need to raise our process IQ. We need to make our processes more intelligent, and I think that it's a key capability that enables that.

Rudy Kuhn: I truly believe that there are only two ways how business processes can be improved in general or in principle. It's about, you either increase efficiency or reduce risk. And the biggest challenge always is that people simply don't know where to start. We know the processes are not perfect, but where's the pain? Where should we really start? And process mining is a great, great way to show the challenges and then really take action. And I think this is something that we learned over time, that just seeing the x-ray is not good enough to fix the problem or to fix a broken bone. If you break your bone or your foot and you end up in the hospital and they send you home with the pictures only and tell you, "All the best and good luck, get well," this will not solve the problem.

Alexander Rinke: 100%, 100%.

Rudy Kuhn: Do something about it.

Celosphere 2023: The global process mining conference for changemakers

How has process mining changed in the past decade?

Alexander Rinke: I think for me it's the maturity on multiple dimensions, maturity in terms of community. I was at two of our World Tour stops already, you can see that there's a real community around and real adoption. It's not just a few people talking about it, but there's companies that are all in that are deploying this across the board. I think also in terms of not just–moving from understanding your processes to really running it based on these insights in the day-to-day fashion, I think that's been a big change.

Rudy Kuhn: Maybe also a big change is really to see the level of professionalism in our customers. So once people really grasp the value of process mining, they start to work on it. Very focused on building a center of excellence, having a team running process mining across the organization, going from business department to business department, looking at processes end to end.

Alexander Rinke: It's very exciting to work in the space where companies realize this is so transformative that we are actually going to create dedicated organizational units and even change other organizational design patterns because of the power of this technology.

Rudy Kuhn: And the thing about process mining really is, it's one of these technologies that are not 100% business and they are not 100% IT. It's really in between. It's the interface. Without IT, process mining is not possible, because we need the data, we need access to systems. But if you only show the processes to the IT people, they will say, "Okay, it's nice. It's another way of business intelligence maybe." But if you show it to the business people, they will realize the value. So this working at the interface between business and IT, this is actually what I really love about it.

Alexander Rinke: Yes, that's right. That's right.

Why are you excited about object-centric process mining?

Alexander Rinke: So why are you excited about object-centric process mining and also the underlying data model?

Rudy Kuhn: If you look at one process only, let's take, we talk about manufacturing or production. So if you see a production order is running late, of course you are worried because you want to make your customers happy, you want to deliver the goods in time, and you are searching for the root cause. But maybe the root cause is not the production process itself, it's maybe the purchasing process because some required goods are not available and will not be delivered in time. So a challenge or the problem in the purchasing process is actually causing a problem in the manufacturing process. And with object-centric process mining, you can finally see the influence of processes.

So it will give you more insights, it'll really help you to understand where the real root cause is. We can take this maybe even farther, I don't know if we have done this before, but even across organizational boundaries.

Alexander Rinke: 100%.

Rudy Kuhn: Like a supplier and a vendor working together, the combined purchase and order to cash process.

Alexander Rinke: 100%.

Rudy Kuhn: So we have all these interfaces, and interfaces, typically that's the point where processes break or where the issues are. And now combining data from two organizations, from two totally independent companies, in order to visualize, improve the process where they touch, this is again possible with OCPM.

Alexander Rinke: Yeah, 100%. I completely agree with you. And it is interesting, when I reflect on the history of the company, there's been three major product evolutions in the past. There's been the initial inception, the first process mining system. The second was then that we developed our own memory engine to be able to really handle large volumes of data, do it quickly, provide really deep process analytics, custom data processing capabilities for process mining. The third was then to go in the cloud and also embed action and monitoring and data pipelines and all that thing, so it becomes a much more holistic system that can be used for ongoing monitoring, for action initiation, orchestration, as you call it. And now this evolution that effectively allows you to centralize all of your process data in a single structure, which I think is going to ... I'm probably more excited about this version than all the other things in the past.

And I feel a little bit like in the old days where I was running around telling everybody how excited I am about process mining and people were looking at me like, "What's this guy talking?" Now I tell everybody how excited I am about the next version of it and the evolution, because I think it really allows us to become a core part of the enterprise architecture that is a system to really raise the intelligence of your execution and your processes holistically.

So for example, when I talk to people even in an accounts payable function, what they often tell me is that they have a lot of issues because of the upstream process. Like in purchasing and in sourcing, for example, payment terms are not aligned and that's why they have an issue. Whereas in purchasing and sourcing they're not even aware of the downstream impact of their decisions. So they can make better decisions if they, for example, can measure working capital impacts downstream. So I think that it allows an enterprise to have a more intelligent way of operating.

I tell you another user that's going to be really interested in this, it's not the human but the artificial intelligence. If you think about having this understanding of business processes in an organization, this knowledge, and feeding this or exposing this to AI systems, it makes these AI systems really work in the context of the business domain. Because nobody would want to, even if it was possible, no company would want to share all of their data with an LLM. That would be crazy, because there's a lot of uniqueness to how you operate. The prices you have with your suppliers, the specific rules and mechanics that you have in your business processes. But if you can combine the reasoning capabilities of an AI system with the incredibly strong foundation of our platform, you can provide some real value that I think people can't even imagine today.

How does process mining enable end-to-end process orchestration?

Rudy Kuhn: RPA and automation is a great technology, but what is really missing is the overview of the entire end-to-end process. And that's exactly what we can do with process mining, especially with object-centric process mining, looking at all the processes. So just imagine this control tower view on the process. You'll see exactly the different systems, you'll see how the process runs, you'll see the bots working in between, and you can really orchestrate everything. And if you discover that something is wrong in your process, because we are capable of analyzing the data in real time, so you can immediately trigger a workflow, you can trigger an RPA bot, you can do whatever is necessary to fix the problem we just discovered by looking at the process, by looking at the data, and make sure that we basically heal the process as it happens immediately.

Alexander Rinke: 100% agree.

Rudy Kuhn: Yep. Just one last picture here. We often talk about process mining like x-ray. I think that's a great picture, because it really explains that you make visible what was hidden underneath before, you couldn't see. But honestly, I think that process mining should be more like a smart watch. It's continuously monitoring my performance. It's counting my steps, it's measuring my pulse. And if I don't move along or if I don't move enough, if I sit for too long, it'll remind me, "Hey, get up. Take some steps."

Alexander Rinke: It is much more proactive that way, right?

Rudy Kuhn: Yeah.

Alexander Rinke: Yeah, 100%.

Rudy Kuhn: This real time monitoring and real time ...

Alexander Rinke: Observability.

Rudy Kuhn: Observability.

Alexander Rinke: Exactly.

Rudy Kuhn: Actionability.

Alexander Rinke: 100%.

What should companies think about when looking to drive digital transformation and benefit from process mining?

Rudy Kuhn: Digital transformation actually is a big word. And when I look at my past, probably the last 25 years I was constantly working on transformation in some way. So it started with digitization back in the late '90s with process digitization, process improvement, and then of course process mining. The last four years I was working on process automation, and now I feel like it's the next step now to transformation. Transformation and orchestration. And if you think about processes, we always talked about PPT.

Alexander Rinke: Exactly.

Rudy Kuhn: People, process, technology, or process, people, technology. So if your employees are struggling with the process, train them, make them happy, help them do a better job. If your process is not executed well enough, too little efficiency, too much risk, you can change the process. Or you can apply technology, automation. Automation has come long way to liberate people from mindless and boring tasks. But what typically you will see is that companies are combining all together. And in order to know where to start, you need transparency. And what provides transparency?

Alexander Rinke: Process mining.

Rudy Kuhn: Exactly. Process mining is the key.

Alexander Rinke: I agree. I think that digital transformation has so far transformed individual functions and how people work, but it hasn't really completely changed how enterprises operate. And that's clearly the next step. And I think that intelligence, intelligence provided by process mining especially with our enhanced architecture, is a centerpiece to that. And I think it enables organizations to capture, exponential amounts of value. Because what I'm always thinking about when I hear CIOs or customers in general talk about these transformation initiatives, I'm like, "How do you measure the value?" Because it's really important when you're investing huge amounts in cloud transformation and other transformation system upgrades, other transformation initiatives, is how do you actually make sure that there's a real outcome for the business? And I think that we provide a very elegant way.

Even now process mining has become a very established category. There's a magic quadrant and all of these things. So it's there, it works. Many, many large customers have adopted it. And I think that the ability to easily demonstrate and also enrich the value proposition of what you're doing and guide you to where you actually can generate value is a game changer. Because that way I think it has a much bigger impact in organizations. And it also is more relevant on a more senior level. I don't think there's many technologies that can have 10X or even 100X ROI at large scale.

Rudy Kuhn: Yeah, I really think that process mining enables transformation. It enables change. Because again, you see what's really going on and you don't have to guess any more. And you know the quote, without data you are just another person with an opinion. So process mining actually provides you the root cause. It gives you the numbers, you can come back, you can measure the same process, and you can really kick off the transformation process in your organization and really change something for the better and really create value.

Alexander Rinke: Yeah, process mining has come a very, very long way. But when I think about where our architecture is going, where this whole space is going, it will be so much potential. It really becomes a foundational technology that is used everywhere in the company that is becoming just how things are getting done and embedding itself into the fabric of day-to-day execution, to your point also about orchestration. And I think with an architecture that we've developed that enables that, I think that we're just getting started to realize the potential of this technology in the space.

Rudy Kuhn: Yeah, because as you said, most of the customers are only using it now for the core four processes, and there's so much more.

Alexander Rinke: There's so much more, 100%.

What are your predictions for the future of process mining?

Rudy Kuhn: The utilization of AI will transform process mining both on the side of working with the system, also the user interface especially. And for me, the biggest change or the biggest expectation for me is the utilization and the actionability. So from just observing and enabling change, actually in driving change as a control tower overlooking the entire process landscape.

Alexander Rinke: 100%. So I think in terms of prediction, I would offer three. First, I think this expanded architecture with object-centric will become a new standard in the industry because it's so much more powerful. Secondly, we will observe process mining and especially with the object-centric architecture to become not just exposed to people to gather insights, but to orchestrate people, AI models, and any sorts of enterprise software like systems, bots, integration systems. Because then you can really infuse intelligence into the day-to-day execution of the business processes.

Rudy Kuhn: The end to end process, yeah.

Alexander Rinke: And then third, I think customers will get a ton of value from Celonis.

Rudy Kuhn: Absolutely.

Bill Detwiler

Senior Communications Strategist and Editor Celonis Blog

Bill Detwiler is Senior Communications Strategist and Editor of the Celonis blog. He is the former Editor in Chief of TechRepublic, where he hosted the Dynamic Developer podcast and Cracking Open, CNET’s popular online show. Bill is an award-winning journalist, who’s covered the tech industry for more than two decades. Prior to his career in the software industry and tech media, he was an IT professional in the social research and energy industries.