Mining Your Business

20 | Lars Reinkemeyer, Vice President of Customer Transformation at Celonis, introduces the ideas behind Celonis Execution Management System

October 27, 2021 Mining Your Business Episode 20
Mining Your Business
20 | Lars Reinkemeyer, Vice President of Customer Transformation at Celonis, introduces the ideas behind Celonis Execution Management System
Show Notes Transcript

Process mining is shifting from visualization and exploration to proactive approach. Celonis, the leader in the process mining vendor market based on the Gartner PEAK matrix, can offer all of that in its Execution management system - EMS. To tell us more about EMS, we have invited the vice president of Customer Transformation at Celonis, former head of global process mining department at Siemens, guest lecturer at UCSB and Stanford and author of the book Process mining in action, Dr. Lars Reinkemeyer.

00:00

Patrick:

Welcome. Welcome back to the Mining Your Business podcast. Show all about process mining, data science, and advanced business analytics. I am Patrick and with me, as always, my colleague Jakub, hi there.

 

00:11

Jakub:

Hey, Patrick.

 

00:13

Patrick:

Here to tell us all about the Celonis Execution Management System. Lars Reinkemeyer, Former head of global process mining at Siemens and now vice president of Customer Transformation at Celonis. You excited, Jakub?

 

00:24

Jakub:

Hell yeah!

 

00:25

Patrick:

I'm excited, too. Let's do it.

 

00:36

Jakub:

The future of process mining is now. Today you don't only mine, visualise, and analyse the data, but more and more you focus on the execution parts of process mining, building proactive notification and alert systems, planning and simulation of your processes. And of course, automation of your workflows. Celonis, the leader in the process mining vendor market, based on the Gartner Peak metrics, can offer all of that in its Execution Management system, so-called EMS. To tell us more about all of this. We have invited the Vice President of Customer Transformation at Celonis, former head of Global Process Mining Department at Siemens, guest lecturer at UCSB at Stanford and author of the book Process Mining in Action. Dr. Lars Reinkemeyer. Lars, welcome to the Mining Your Business podcast. It's an honour to have you on the show.

 

01:28

Lars Reinkemeyer:

Jakub, thank you so much for inviting me and really happy to be here and the future, which is now and sharing a bit of my experiences and my vision about process and execution management.

 

01:38

Jakub:

Yeah, I'm definitely persuaded and sure that this will be a great episode for our audience. As Processand as a platinum implementation partner with Celonis, we want to stay at the top of our game. We have to honestly, if we want to keep up the pace with the best. Luckily, EMS is giving us much needed tools to keep the edge and always deliver our customers the value that they are looking for.

 

02:02

Jakub:

The number of changes and improvements that have been introduced in the Celonis environment in the last couple of years is really staggering, and it all really started with the much expected move to the cloud, and it never really stopped there ever since. Essentially, each new feature exponentially increases the number of ways you as an organisation can leverage the Celonis tool and process mining in general. But before we move into what Celonis offers now, I would like to step back a little and look into what Celonis environment was offering in, let's say year 2014, the year you have become the head of Global Process Mining at Siemens. So, Lars, looking back in 2014, what was even your perception of process mining?

 

02:46

Lars Reinkemeyer:

Yeah, good question Jakub. If you jump on the time machine and zoom back in 2014 process mining was completely new, it was completely unknown. To me it was something which I thought, OK, what is this about? You know, and by the time I was in charge of business intelligence at Siemens, we had established a huge data lake where we used extensively data analytics what you would call today Tableau, Qlik View, you know, making these kind of reportings. But I felt intrigued by the unique value proposition of process mining in the sense of understanding business processes. And when we started in 2014, we used it intensely as a kind of x ray to get an understanding about actual processes, you know, not the way how you believe processes are happening and how you have modelled your process but really understanding how does my process happen in everyday life. And that was for us an amazing eye opener. When we started looking into logistic processes, procurement processes and understanding the real complexity, the real activities and the real challenges why those processes would not be performant as they were expected to be performant. And this was the starting point for us, you know, back then to use process mining and get this kind of discovery or switch the light on on the processes and to be able then to understand what are the pain points, what are the reasons for delays, what are the effort drivers? Why does the organisation not work at full performance or full capacity since there are, you know, delays where you have purchase orders for a simple mouse that have ten approval steps with people in this process don't know of each other and where process mining allows to understand, oh, there is so much inefficiency since you have so many people approving the step there. And that was for us in a huge eye opener which we started 2014 and then for several years used it across different processes in the global teams organisation.

 

05:01

Patrick:

Now when you first saw process mining in action, did you have some sort of clear problem in mind or challenge that you were trying to solve with it, or was it more an explorative venture where you were seeing, hey, what can this technology do and what can we find out about our own internal processes with it?

 

05:22

Lars Reinkemeyer:

I think it was more of a learning journey which we had there since, you know, I was in corporate IT, we had this technology and we all were excited by technology and we thought if we have a hammer, every problem is a nail. And, you know, we went around and tried to tease everybody with that. But it was a time I understood that the more X-rays I did and when not really there was an impact, I understood that I have to understand first the pain point from the problem from the business process owner, you know, understand the problem. So with the learning, we started talking with people from procurement and logistics saying tell me what is your pain point? What in this process has been the topic which you always wanted to understand? Where did you want to get transparency? And they coached us saying Well, it would be great if we could understand across hundred thousand of suppliers, if we could understand duplicate payments, if we could understand in the process the deviation between the payment term on the purchase order, all the confirmation, if we could understand maverick buying in a purchase process. So they taught us, the process experts taught us their typical processes where they assumed there were inefficiencies and thus allowed us to calibrate our x ray to allow them to find exactly the error drivers, the inefficiencies and then start action on that. So it was a joint learning which we had to calibrate our technology to these specific process requirements and to help the business owners to make their processes more efficient.

 

06:56

Jakub:

I will slowly kind of want to try to transition into this EMS topic. But before we do, I guess also, as you were in Siemens, you faced a lot of challenges and I can assume that some of them will revolve around the principles that you laid down in your book, so-called three P's, The Purpose, People and Process traces. But I will actually ask a question that I have to say, there was a comment under one of your LinkedIn posts where a person asks, Where did you fail as an organisation?

 

07:31

Lars Reinkemeyer:

Right. And you know, it's always an exciting question where however you know, you tend to be cautious saying, OK, what failures did we have there since some people are a little bit like getting audited, but you know, there's a couple of things which, which I'd rather call it a key learnings which we made there. You know, like the first part, first of all saying it's not about the technology, it's about the people, it's really about understanding the people, how they work and taking them along, getting them engaged and making sure that with a decent change management, you excite people about this new technology, new possibilities, you know, when you want to guide them towards automation, people are hesitant, reluctant, and when we started, we failed in understanding this necessity to take the people along and, you know, get them engaged and get them excited and get them on board. So this is the key learning. Maybe, well, another topic, which was probably is a funny story, which where we failed, we actually failed and I can admit that is that, back in 2016 our CEO was complaining about the complexity in the organisation and I said hey this is exactly the topic which we can provide facts and data for measuring complexity in an organisation, you know, and the CEO at this whole programme saying let's tackle on this complexity. And I said I have a complexity monitor where it can measure exactly the complexity in each of the organisational units across Siemens and compare them and make benchmarking and best practises and everything. I failed to be honest and so somehow I didn't get the buy in from and from the board. I mean we spoke to a board member, proposed it, but then it didn't get the right attention on that to be scaled across the global organisation. But I still believe that was a great idea. But we feel by the time that it was a great idea to have a kind of complexity monitor which allows an organisation to get transparency across the efficiency in the different departments, make benchmarking and then obviously drive action based on that. So these are two examples where probably we failed in a sense, you know, and there was a whole learning journey there also failed sometimes to get access to data, since the idea of process mining is obviously to have all these event logs and sometimes we've failed since our customers said, hey, can we have a huge combination of data from transactional systems, but also from Outlook and Excel and you know, where you cannot get any digital traces. And there we failed on a couple of requests though today technology allows much more to get the different sources and also with task mining being able to record what people are doing. So in that failure, we probably improved also with our technology and moving towards EMS and task mining. Today we are much more capable than we would have been back in 2014.

 

10:29

Jakub:

Yeah. So first of all it's very comforting to hear that you in your organisation and in your experience were facing similar issues that we are facing on a daily basis with our customers. It's, you know, it's always the same, it belongs to the work, it belongs to the project and it belongs through process mining. However, process mining in 2015 and even still in 2018 was different than what it is now. And you changed companies. Now you actually work in Celonis as a leading organisational transformation. And I would like to know what is your job really now? What are you doing on your daily tasks?

 

11:12

Lars Reinkemeyer:

Yeah. And before I answer this maybe on the first point there, how the evolution has gone and what's been also a key reason for me to change, you know since then what I see when we started was process mining was very much getting an understanding about processes and efficiency. So in the sense of discovery inside looking back and understanding and then applying human intelligence to derive measures and activities, you know, and Celonis has evolved with execution management system towards a more proactive, forward looking approach, you know, which I found so compelling in the sense of saying, let's replace the human intelligence with artificial intelligence, let the system read the data and then suggest and propose to the people what they should be doing. And this is the whole idea of execution management systems to enable organisations by intelligently reading their data and then suggesting where they should get active and enabling them with execution apps to do these parts, you know, and this I think is a tremendous, a huge step forward, which for me was one of the key reasons as a senior executive of Siemens to resign and join this exciting company Celonis. At Celonis I'm in the service delivery and as a VP customer transformation, my mission is to bring in my experience, which I gained the last couple of years here, and advise companies on their transformational journey and what this means is basically that on a daily basis, I'm working with companies supporting them in understanding how they can apply our execution management technology, but also what it takes to drive organisational change. And this is not only about technology, but this is also about change management, this is about operating models, this is about engaging the people, this is about building centre of excellence. And this is where I and my team have a strong focus bringing in experience and methodology and coaching companies on using the technology successfully, but also driving process transformation across organisations.

 

13:20

Jakub:

I guess the challenge now isn't really the mining or the insights from process mining, but it's really the action and this is also a topic that Celonis now is really trying to address. And I'm just wondering, knowing what you know now and having the tools available that we have now, would there be some topics or some problems that you had back in your Siemens career that you would be able now to address much better with the tools that you have at your hand?

 

13:50

Lars Reinkemeyer:

Oh well, yeah, definitely. I think on that I could talk half an hour. I mean, there's been so many challenges starting from the technology side, where we spend a lot of time and effort in finding the event logs in those systems, which today with the standard connectors is so much easier, it's still effort but it has been much, much easier now to connect to the whole range of transactional systems. You know, like not only SAP, Oracle, Salesforce, Koopa, whatever you name it there. This is a wealth of technology which has evolved there. Then the whole part of task mining, you know, being capable to record activities in a call centre and see what people are doing. So activities which are not in transactional systems but where they do something in email outlook, explorer or whatever and record that understand that and combine those processes with the transactional data on the whole part of artificial intelligence and standardisation, you know, is execution apps. Where we have centred solutions on helping companies to process their purchase orders or customer orders in a much more efficient manner, which we were dreaming off a couple of years ago. Today it's reality, you know, having a technology which understands the purchase orders and then provides the users only with the information which are relevant, you know, so it helps companies and the people in their daily efficiency by pre filtering the relevant information and feeding them with the right priorities and things to be done. If it comes to action flows, proactively alerting people on those things and presenting on a still a tablet, everything which needs to be done. So there's a whole huge range. And I mean, to be honest, it's amazing to see this technology evolution and it's exciting looking forward. You know, how it's just to anticipate what's still to come there.

 

15:42

Patrick:

So for the listeners out there that aren't familiar with the EMS, can you give them a little bit of an overview, a small little introduction, what it is, the high level concept behind it and why it's beneficial to them? I know you've given some examples of the things that can do, but just in more of a general sense, give them an understanding of what the EMS is.

 

16:01

Lars Reinkemeyer:

Yeah, absolutely. And I think the general understanding is that Celonis has grown ten years with process mining. You know, this x-ray of processes and this is our foundation, this is our asset, where we also are unique in the sense of being able to do this for huge companies with hundreds of millions of activities and visualisation of hundreds of thousands of different process variants. But this, again, is the foundation. The execution management system now is building on that. It's the next step forward in the sense of having a focus on helping companies execute their processes smarter and faster and more efficiently. And this whole idea saying building on process mining on an then thorough understanding of how the process is actually happening, but then was an execution management system being able with real time data ingestion, getting that information, aggregating this information, and then providing the users with those topics and priorities which are to be done. So what does it mean in reality? Somebody from procurement does not need to look into process mining and looking to thousands of purchase orders, seeing which are going to be delivered too late or which are going to be paid too late. But the execution management system is capable to detect this from the actual purchase orders and then present the users only with those activities where they need to look into it. Since this is a purchase order which is to be paid today to receive the discount or this is a purchase order which according to the current activities which we're seeing in the system, is going to be probably too late. So an alert the user saying, Hey, this is where I should be looking into since this delivery will be delayed and you might want to think about urgent actions to do something. So the whole idea of execution management system is to enable users in a smart manner with standart applications and with action flows and proactive support to do their job in a much smarter way. And then also to build on machine learning, artificial intelligence and predict based on the process mining data which we have, what is going to happen, predict which deliveries going to be delayed and then execute this part. And the vision which we have there is at this point in time we're supporting users. But then also in order to execute certain things where the system sees, oh, this is going to be happening too late, let me also execute this fast delivery for example, you know, or it may remove this delivery block. And this is where the technology gets smarter and smarter and this the idea of execution management system.

 

18:33

Jakub:

So the EMS is really divided into couple of parts. And this is something I looked at on Celonis website and basically there's now these modules and those are real time data ingestion process and task mining, blending and simulation, visual and daily management and of course, action flows. Now, let's take a look a little at each of these parts. So you already mentioned this real time data ingestions, which really means that you can get the data from your entire process ecosystem into Celonis. You said that there have been already some development that, you know, compared to the history, the data transfers are simply way easier. There are so many prebuilt modules on how to extract them. And then why is it so important or is it necessary to do these real time extractions now? And how does it help an organisation to have the data almost on life basis compared to let's say on a weekly basis or on a daily basis when you know, you can still think of process mining as a business intelligence tool to a certain degree.

 

19:44

Lars Reinkemeyer:

Yeah. And then it goes far beyond business intelligence. You know, this is what we want to do is not only kind of reporting but we want to enable the organisations to execute things smarter. So the idea about real time data ingestion here is that we provide a capability to take those digital traces from any source system virtually in real time and thus enable organisations to do certain activities in a smarter manner. For example, this is relevant if you are in a manufacturing environment and we were using Celonis in a couple of Siemens plants and there the necessity is with real time data ingestion to see where in my production chain is the bottleneck since there is, you know, something in the whole flow is a bottleneck coming up, getting this real time data alerting and then being able to take mitigating action based on that. Another example where real time data ingestion is very relevant, is in call centers, if you have call centres there are reactive alerting. Like Dell is doing this alerting the agents their real time what's happening where the delivery is delayed, what payment is coming up. So to have a screen a 360 degree view on the customer and also then some kind of prediction what's going to happen so that the consultant can support the customer real time on the phone in a much more profound manner. And this is where we see the whole real time data ingestion evolving.

 

21:14

Patrick:

Now when you say real time what kind of timeframe are we talking about? Because you know the usual method is getting the fresh new data from that day during the night or something, building the reports and having it displayed the following day if you're lucky. So if you're talking real time data ingestion, how quickly could it be from the data point that has the bottleneck to somebody figuring out that something needs to be done about it?

 

21:36

Lars Reinkemeyer:

Absolutely, Patrick. And I think this is a very good question since obviously there's a whole chain in between, you know, and since like we also looked at the latency of data replication from the ERP systems in Asia, for example, where you have a natural latency from the system, then the transmission, then the preparation and then the visualisation. So it's a whole chain which brings in certain latency. And to be honest, I think what always needs to be considered is for the use case, what's the benefit of real time data availability? Since if people look into their daily purchase orders or customer orders, you might not need real time and daily update is sufficient. And to be honest, in my experience, the majority of use cases which I know so far I'm happy with having daily data updates for taking decisions and process improvements. Now the part about real time is really now that we are moving towards a real time where you have less than a second of a latency on showing the data and using the data. And as I said on those two samples, with manufacturing and call centres, those are also environments where this is necessary, where there are specific use cases where the agent needs to have that data immediately on the fingertip to be able to advise the customer on the call on the phone there. And that's why we are heavily investing into this real time replication. Also looking into Kafka technology and enhancing our capabilities with there will be a big announcement soon as we're taking a major step here to become more powerful on providing that real time replication and information.

 

23:18

Patrick:

So what you're saying is that somebody could get an alert about, hey, there's something wrong on your assembly line, they could walk out the door and see it for themselves in real time happening in front of them.

 

23:27

Lars Reinkemeyer:

Exactly. That's the idea in the manufacturing, to see an assembly line real time. There is a bottleneck and there is material stocking piling up, you know, and this is real time alerting with push alert, then being sent to the person who can go down and remediate that problem.

 

23:43

Patrick:

Well, so I'm going on to the next topic, process and task mining now. How does this benefit a organisation figuring out the small details that aren't transactional? How does that benefit an enterprise?

 

23:58

Lars Reinkemeyer:

Well, I think the combination there and the power is to bring together the process mining, which is data from transactional systems like ERP systems and task mining, which is data which you record like for example in a call centre. So the idea of task mining is to say recording how does an agent in the call centre process a purchase order where the person might be active in the ERP system, in Outlook, in Excel, in something else, you know, across different tools and getting that kind of snapshot of this process. And the power is to combine the data saying if I want to understand the purchase order processing across my whole system, I not only want to see what's happening in my ERP transaction system, but I also want to see what is a person hands on doing. And the power here is to combine that kind of process information, these event logs from process mining and task mining understand thoroughly the whole process and then identify again where do I need to take measures or also in respect of benchmarking, saying why is a purchase order being processed faster in a shared service centre in one region compared to another region? What can we learn from the single activities? What best practises do we have and how can we can coach and support our service colleagues to be more efficient? And this whole idea of combining process mining and task mining into one complete understanding of the actual processes a company is executing.

 

25:34

Jakub:

So in Celonis always the main feature, the main centre was the Process Explorer, basically that was the main thing why companies were purchasing Celonis in their organisations, to be able to visualise processes. Now with all those features that are coming up, isn't this somehow leaving this process behind while now you can basically have these actions, notifications and essentially there could be users that don't even that won't even need to use the process explorer. And if this is the case, how do you still leave the focus on there or is it even intended?

 

26:16

Lars Reinkemeyer:

I think the process explorer is a very powerful tool and people in the audience who might not be familiar with process explorer, the process explorer basically is the x ray image, you know, where you can see in this process explorer, the actual process flows, the actual process variance, the actual process times and I think this is very interesting for data analysts who want to understand what's happening, see the process variance, see the root causes for delay. You know, but in my experience, the biggest impact and the biggest number of users is those people who daily have to execute purchase orders, customer orders, logistic processes. And they often look back at Siemens. Out of these 6000 people who are in our use, a community minority was using actively the process explorer for going deeply in understanding the root causes. The majority was using more of dashboards where they could then get the right information, the fingertips and start working with that. So what does it mean? From the process explorer we would see which are the customers, which are sending paper and fax orders, and the people in automation would get a report where they can see, OK, these are the customers whom they need to talk to since they cost enough manual effort and rework, you know? So it's this actionable insight, which is the important part there, on which we focus on, and that's where we build on the Process Explorer. But rather present actual insights to the users plus and also enable them with execution ready steps in a smart and easier manner.

 

27:52

Patrick:

So moving on to the planning and simulation part now, I was very, very interested about this when I heard about it recently, the planning and simulation, for the listeners, it kind of lets you model and simulate future state processes to understand the impact of decisions. So can you give us a little tease as to what it really does and what are the what if questions inside the EMS?

 

28:14

Lars Reinkemeyer:

The whole idea here about planning simulation is to say if we would have an ideal process which really allows companies to execute at maximum capacity, what would that look like and how far away are we from that? You know, and the approach would be taken here is a bit different to traditional business process management companies. You know, there are other companies which have a strong back hold on BPM where they model the processes and you know, said this is how the process supposed to be. And that's been the core focus. I don't really believe in that part. I believe rather in the approach saying, let me understand the actual way a process happens, then let me think about how the process ideally should be looking like. Then let me do a kind of conformance checking here and understand where my biggest deviations through my ideal process are and then take action to improve my process. So the idea about this whole concept of planning and simulation is to plan the ideal way how a process should be, but also with a clear understanding of matching the real life with this ideal planning and this simulation and then see how do I get or what kind of measure do I take to get to that ideal flow, you know, and to have a kind of conformance. So what we will be providing here is the capability for companies to say, OK, currently my execution capacity is at 50% on time delivery. If my planning I want to go to 100% and simulate that you know, how would the process be looking like and then say, how do I migrate? How do I transferred to get to that vision? This is the whole idea of planning and simulation.

 

30:02

Jakub:

How can you predict the consequences, I wonder, is it just based on the historical data or how do you then are inserted into what is happening and predicting really the future?

 

30:14

Lars Reinkemeyer:

Well, predicting I mean, predictions more about saying, here on the simulation part is more about simulating the way how things are happening and then predict, if my purchase order would always happen in that kind of form with that kind of process flow, how much time would I save? How much effort will save, how much more automation would I have, how much more customer satisfaction would I have if I could deliver more on time? And that's the kind of prediction and scenario which we providing to simulate saying this is where we want to get to in order to achieve maximum capacity. But then also looking and comparing, saying, what's my current status quo? And then obviously the important part is always on defining actions, saying how do I get to my future state of efficiency there and how can I track that? If I'm on the right on track there and improve my capacity on my execution performance.

 

31:13

Jakub:

In your book, you are mentioning that the conformance of the processes is somehow an under-utilised feature in general in process mining. And I have to say I can confirm this as I also don't see many customers who are really looking more into conformance, but really more on this action part of process mining. Do you think that this whole planning and simulation part in EMS can maybe put a bit more focus back on the conformance?

 

31:42

Lars Reinkemeyer:

Absolutely. That's clearly the intention here, Jakub, to have a stronger focus on basically what conformance does is to bridge the gap between the simulation and the reality. I think this is the power which you have there and saying, OK, let me simulate my ideal purchase order process. Then let me with Celonis to identify my actual way how my purchase orders are being processed. And then let me see on the conformance where are my deviations? And this is the simple approach. There was a conformance checking to have a step by step approach, saying, let's see every single deviation in the actual process versus the target process. Where do I have a nonconforming process? And the way how it's been in use in several use cases is then to go through and say, OK, there are certain non-conforming activities which are whitelist, which don't hurt, you know, which don't have a bad impact. And then there are other non-conforming activities which are blacklisted since they cause rework, inefficiency or whatever. So with this conformance checking, you can flag this out and then start getting active and remediating those blacklisted nonconforming activities.

 

33:01

Patrick:

Do you see a barrier for a lot of users to get on board with a simulation of their processes when they suddenly numbers and KPIs appear what would be in a year or a month's time from now, seemingly from like a crystal ball sort of algorithm that figure this out, you think there's some sort of barrier that would keep people from believing it? Or do you see it differently?

 

33:26

Lars Reinkemeyer:

Well, first I see different user types. So the majority of our target group and users are those people on a daily basis who work with purchase orders and use our EMS technology execution apps to execute just their standard purchase orders. Those as a core people, core focus here. Now the barrier and the target group for conformance are more the people who are interested in understanding the process analytics. And this is where the power comes in for somebody who in the past was BPM has been on the dry dock, just modelling processes and thinking this is the way how reality is. And now with the simulation conformance checking we enable them to match the reality, what the reality looks like, you know, and the barrier I mean I was surprised that to be honest I tried to engage and excite many people there who are coming from the BPM side and many of them didn't show interest, you know, which was for me and maybe I failed there, you know, the failure which I had in my past, you know, and some of them just told me Lars, you know, my task, my job description is just about process modelling. I just need to do price modelling. And I said okay, but wouldn't it be interesting to see reality? And they said no, I just need to do documentation modelling. Ok, so I failed to evangelise, you know, others who have a different mindset and who see this power and potential saying, hey, this is so exciting to see the reality matched versus my BPM model. I think this is where there is a huge, huge potential where they can use that and do this conformance checking and there's some people. But again, in relation, if we look at the users, I mean this is rather a smaller focus group of process experts who understand the power of having real time data match with model data, but the vast number of users is definitely across different applications.

 

35:27

Jakub:

So as a next part of this EMS environment or ecosystem, we've got a visual and daily management and in here what we are supposed to get is to have ourselves delivered some proactive insights, prioritise tasks and so on. And I imagine that back in the days, if we say like this, when process mining was a bit younger, we had a data analyst who was doing some work in the reports and to get some insights, he or she, the data analyst, to understand what they see and you know, click through the report and find the potential for improvements. So now I imagine that Celonis is trying to go in a way to actually give them these insights somehow, automatically. Could you walk us a bit through of more the thoughts behind it?

 

36:23

Lars Reinkemeyer:

This yeah, absolutely. And this is exactly where the direction which you indicated there with this visual and daily management, the intention is to bring in all our experience of ten years, learning from people in procurement, sales, manufacturing, logistics, on how they work on a daily basis and providing them the right applications, the right data, the right inside where they can immediately derive action based on that. You know, so really providing to hundreds of people in procurement a screen where they can see those relevant purchase orders, which they need to act on right now to see based on the process where our blocking points and processing of purchase orders which are to be removed and bring that up to the attention of the user. So here we're looking into a wealth of different standard applications and insights which are used by our customers on a daily basis where they can then immediately derive action to take give. Another example, we have the execution apps for order management where the app enables the people then to show them which customer orders are to be looked at, which are going to be delivered too late, so really providing actionable insights in a standard dashboard in a standard app. And this is where also our apps got much smarter since in the beginning when I started we customised these dashboards ourselves, you know, we try to understand, OK, what does the person need for daily work? And then we try to provide that right information. Today was all that experience which Celonis has gained throughout the years. This has been provided as a standard since we know how a company executes on certain activities and this is what is providing incentive to customised dashboards and easy to use across hundreds of people in an organisation.

 

38:21

Jakub:

So no more bookmarks.

 

38:24

Lars Reinkemeyer:

Yeah, and I mean the art is always, you know, to have the right mix between standardisation and customisation. So that's obviously also possible but to have the biggest impact, you know, and to be honest I always wondered why are there thousands of people processing manually purchase orders in an organisation you know, why does Amazon do this all fully automatically? So why not standardise the way how commodity materials are purchased that should be standard and easy and not with individual bookmarks and everything, but obviously the technology is both past standardisation or customisation depending on where the customer wants to go.

 

39:08

Patrick:

So getting from these insights for these individual users to action flows. Now this obviously comes in part with the acquisition from Celonis of Integromat, I believe, implementing these action flows, these actionable insights in a chain of actions to be done by either the system or the person. So can you kind of walk us through what this means for an enterprise and what Celonis does that is so magical?

 

39:37

Lars Reinkemeyer:

Yeah, and this is the whole exciting part there about, you know, building a virtual companion, which has a certain intelligence to support people and what they do. So the idea of action flow is to have a proactive support for the users. And we started a couple of years ago with the so called Action Engine, which was customisable, to send alerts to the user saying this is the delivery which has the highest priority right now. And initially sending an email and providing the information on the silver tablet. Today, this is done in a more intelligent form in the apps where the users get an alert saying this is the top ten deliveries, which you should prioritize and which you should execute. If you want to execute, just click here and then the system would write back into the transactional system and execute whatever measures taken like an express delivery or customer information. So making it easier for the users to do their daily work. Now the action flow is built on the Integromat, as you mentioned, the acquisition, which we did, a company which is very strongly focussed on having no code customisation for individuals processes and flows. So what does that mean? What we enable our people, our customers here is that they have a very easy drag and drop interface where they can build their own action flows and processes. So for example, if they say I want to have for my purchase order processing, I want to get an alert from my ERP system, then I want to be sent an email, I want to have that documented in an Excel file and then I want to take another action inform a colleague, for example, this is a typical action flow which a user can do by simply dragging and dropping those different input fields, you know, like your ERP data and outlook data and Excel data into a visual flow and thus build own action flows and customise them depending on what they want to do. And the whole idea is to compliment people that they don't need to open the ERP and the Outlook and the Excel and then send something, you know, but have the technology support them in doing that in an easier manner. In a smarter manner. And this is where, you know, I think this whole the exciting part is in understanding how people are working in enabling them to do their work in a in a semi-automated manner. And then also in the future, have the intelligence in alerting them or even auto executing things. This is where our journey is moving towards.

 

42:25

Jakub:

One of the features that action flow and previously, as you are saying, the action engine offers is writing information back into the ERP, into the source system. Now, this is something that I have a very hard time of coming to a customer and tell them, OK, so listen, we have this process mining tool, great thing. And what we can do is actually write back information into your ERP system that they are basically guarding heavily there already. Sometimes there are big struggles even to ask them to extract the data in the first place. And now how do you change their mind about even writing back into the system?

 

43:03

Lars Reinkemeyer:

Yeah, Jakub, you're right. It's not it's not the first step you want to start with talking to your customer. You know, it's definitely something where it's the more elaborate approach. And the idea, I think it's a huge potential. The idea here is to provide an execution layer on top of the transactional systems, to provide an execution layer where the user in the Celonis interface gets those relevant alerts out of multiple transactional systems and then can just click and say, I want to expedite this delivery and right back into the ERP system to make it happen and initiate the shipment via the courier and freight forward. You know, this is the whole idea, and I'm sure I'm convinced that this is the future. You know, since the power here is to have one single user interface across multiple legacy systems where the users can execute things in a much smarter way. So this is definitely the future to come and where it's moving towards. Now reality, as you rightly say, is a little bit more hesitant. You know, like we've been working with many customers first of all to get an access to the event logs through the transaction systems, you know, which is already a big barrier to convince them that if you stick a connector to their system, there is no impact on stability performance whatever, which has been proven thousands of times. But it needs to be proven every single time again. So this is the first challenge which you encounter there. And then you build the trust and confidence, you know, where people see it. At Siemens, we had a hundred systems connected and it worked smoothly and people from shared service and they told us OK, the impact is like less than 1% on performance, though, you replicate thousands or millions of activities per hour, so this where you build the trust and then once you have that at certain time, then the typically the I.T. guys also open and say, OK, I understand this concept of a transactional layer which needs to right back into my source system. And let's try this. And there's a couple of customers right now where we piloting this, where we testing this and where it works well. And I think this is definitely the way how are we moving on and where the future is to have set the execution layer on top where people get active, initiate action there, and then write back into the legacy systems to execute the actions.

 

45:27

Patrick:

Now, do you think there will ever be or rather soon to be a time where these actions will be defined and be fully overtaken by the system itself so that there's no more user intervention that has to click on any buttons, which to me sounds a little futuristic to say like, some bot triggered something in this system and they get sent to this system and then this bot automatically executes something else. And the difference is that the chain of automation there is a very, very high.

 

45:57

Lars Reinkemeyer:

Absolutely. Patrick. this is the future to go. You know, in my book in my third chapter, I try to give an outlook where Wil van der Aalst looked at it from an academic perspective. And I try to describe a digital enabled organisation where the idea is to have like eighty plus percent of all commodity activities done by the system with intelligence. And today if you look at companies like Amazon, they auto execute the majority of one of those things which traditional companies have a lot of manual effort to do, you know, like processing customer orders, purchase orders, logistics, deliveries, you wouldn't think that Amazon has got tens of thousands people processing the order which you place online there, you know, but you wonder then why does why do so many other companies have thousands of people doing this manually? And this is an evolution which we are seeing where I believe that in the first step, all those horizontal core processes like accounts payable to receivables, this will be automated. There will be intelligence like the execution management system and execution apps which will take over more and more of those activities, auto execute this. And only you'll have an exception-based interaction from a human being looking into saying, oh, this is something which I need to look into. This is something popping up here, but this will be less and less. And which doesn't mean that there's no future for humans. I mean, humans will have a different role, you know, in customising, calibrating this technology putting into place doing intelligent diagnostics. So it's shifting, it's shifting the work. But the mundane task where somebody clicks on the times per day on processing a purchase order. I think this is not a future of work.

 

47:47

Jakub:

In your book, you lay down the foundation of something you call a digital enabled organisation. Could you tell us what do you mean by that? And most importantly, let's say that we have an organisation that is, let's say, this lower end of the digital transformation and what it can do to reach that goal to become a digital enabled organisation.

 

48:13

Lars Reinkemeyer:

Then I think this is a whole evolution of organisation which I have to go through, you know, and what I call the digital transformation or customer transformation where I also, in my role as VP of Transformation, I've a strong focus on that understanding. Where is the company today in respect of automation and process efficiency. And then seeing how can we support them in moving towards this digital enabled organisation. So looking into companies which have a high number of manual activities on customer purchase order processing, you know, and then thinking about how we can identify the root causes for many activities, how we can apply automation with RPA batch processing process, improvement of execution apps and then coach them towards a digital enabled organisation and the digital enabled organisation in essence is the idea to say that an organisation leverages the power of digital and automation to a maximum. So to really use execution management systems, action flows, use process automation but also use maybe batch processing, RPAs, so all the technology and have all the exception based human interaction. That is the idea of a digital enabled organisation where many of these centred processes are done really digitally and automated.

 

49:39

Patrick:

In your book you talk about the next big things to happen to the space of process mining. Now we've also touched on this a little bit with predictive and proactive analysis, but would you mind sharing what you think are the biggest things to come in the future in the short term, mid-term and the long term?

 

49:57

Lars Reinkemeyer:

Yeah, I think, I think short term it's and when I wrote the book last year, it was really anticipation, but already now we have many things like proactive alerting and predictive alerting. So these action flows, which we have right now, these are things which already are available right now. So I saw this happening as an enablement. Mid-term what I see there as the big thing is this whole cloud evolution, you know, like all companies now understand the power of moving to the cloud moving from on premises to the cloud. And I think the cloud is not only a technical advantage of having a software as a service, but it provides vast opportunities. What do I mean by that? Just imagine that we're getting Celonis more and more of our customers processing the data in the cloud and what we're building now is a wealth of knowledge about, for example, benchmarking. You know, so looking into the performance KPIs from different companies on standard processes and benchmarking them in an anonymus manner, but still comparing, saying, OK, how efficient is one company on purchase order processing versus other company and what are the secret sources there, why the one company is more efficient. So I think this whole cloud offers huge potential on benchmarking and best practise sharing, knowledge sharing that that's a big thing to come where we see a lot of potential. Third sample to briefly touch on is that I believe there is a huge potential for cross company transparency. You know, just imagine if you take the supply chain of a customer and supplier where you have the event logs where you can understand the actual processes and optimise the processes by driving automation, by driving process efficiency across company. I think that's a huge potential where also we have a couple of customers who are very interested in building that with the data in the cloud, it's easy to take the event logs from both companies, put them together and then see identify root causes and the potential for improvement. So that is another big thing which is which is going to come.

 

52:14

Jakub:

You have me really pumped up for the intermediate future. This sounds all a very exciting thing, but listening to this, there will be a lot of automation, there will be a lot of standardisation. And I have to ask because being a consultant, how do we consultants come into this picture when eventually a lot of things could be even automated and then, well, less job for us.

 

52:40

Lars Reinkemeyer:

I think the status of having a majority of companies 80% automated, I will not see that maybe you'll see that before you retire so but you should be save as consultant was having sufficient work on coaching companies, since to be honest, what technology can do to get this adopted in organisations takes some time. This is just the nature of business this whole transformation and you talk about visions, you talk about ideas and potentials but then to bring it into an organisation, getting this whole transformation and change management in place, This is a business a journey which takes some time. This is where consultants like you are urgently required to bridge that gap, explain to the organisation what they can do, advise them how they can do that, get their organisations adopted. So I see a bright future for the time now and I wouldn't be concerned.

 

53:38

Patrick:

Also Jakub, I must say when Lars mentioned the event log across multiple companies. My ears were perked. I know that sounds like a lot of work and a lot of enabling and lot of work let's say. So I think we're, I think we're fine.

 

53:55

Jakub:

I see a lot of migrations.

 

53:57

Lars Reinkemeyer:

Yeah, exactly. To be safe, yeah, but If you are on that point, if you think about it, you know the potential like every company now is talking about supply chain resilience and supply chain resilience affects not only your own company but also many other companies in front of you and behind you in upstream and downstream and if you take that kind of event log and execution management across those companies, just imagine about the potential which you have there. And this is something which will evolve. I mean, companies will need to be open and to trust each other, but there's huge potential in that. And if consultants like you just steer that idea, you know, and discuss that idea, I think that's already interesting. And then you'll find first companies saying, OK, maybe I as Skoda, if I support Volkswagen, which is my own group there and build the bridges there, it could be a starting point, you know, so I think there's so much coming up. So yeah, bright future.

 

54:56

Jakub:

I have to say that being in the frontline and seeing how these technologies are being implemented and adopted, especially there's still a lot of work to do.

 

55:08

Lars Reinkemeyer:

Absolutely.

 

55:09

Jakub:

Anyway, Lars, I really loved the discussion and I have to repeat what I said at the beginning that the future is now already, and it's really exciting to be there and to see things happening and to transform the organisation and have an EMS of Celonis backing us up on this journey. It's been incredible, Lars, thank you very much for, for joining our little talk here. I know that this will bring a lot of value to all our listeners and you, our dear listeners. Thank you for listening. I hope you enjoy our show, Mining Your Business and if you have any questions, just reach out to us on miningyourbusinesspodcast@gmail.com. We are also very active on LinkedIn and we would love to hear from you if you have any ideas for a future guest. If you would like to hear more stuff about specific topic, just hit us up and we will be happy to share these with you. Lars, Patrick, thank you very much. Talk to you later.

 

56:06

Lars Reinkemeyer:

Thank you, Jakub, Patrick, this was a pleasure, and to thanks for hosting me in your podcast here.

 

56:11

Jakub:

Thank you. Bye bye.

 

56:13

Lars Reinkemeyer:

Bye bye.