The topic “Process” seems to have a renaissance these days. A lot of younger folks are deciding to work in this industry, and a not so small part of this might be caused by the exciting new technology of Process Mining, which seems to be natural for people who grew up with technology in all areas of their life.

Jakub Dvořák

In today’s podcast we are speaking with Jakub Dvořák, who fits right into this profile, and is working at a Process Mining consulting firm in Europe that is currently expanding into the United States.

Jakub is a data scientist during the day, and a financial blogger at night. As part of his work, he is running the “Mining Your Business” podcast (in English) and also the Rozbité prasátko – “Break the piggy bank” – podcast (in Czech).

We are talking about the following topics:

  • Jakub’s background and interests (hint: he’s running two podcasts, and writes a book!)
  • Process&, the company he works in, and their services for process mining
  • How clients use process mining
  • Jakub’s podcast “Mining Your Business” and the renaissance of Business Process Management – data science as the new thing that is interesting for (young) people
  • How do process mining projects work
  • Process mining analysis – what to look for
  • What should you prepare before you start a process mining project?
  • Lessons learned and how to get into process mining

You can find Jakub’s profile on LinkedIn here, and also his and Patrick’s podcast: https://www.miningyourbusinesspodcast.com/ and on LinkedIn (give them a follow, you will enjoy their conversations and guests). via LinkedIn here:

Please reach out to us by either sending an email to hello@whatsyourbaseline.com or leaving us a voice message by clicking here.

Ep. 21 – Consulting in Process Mining: Jakub Dvořák – What's Your Baseline? Enterprise Architecture & Business Process Management Demystified

Additional information

  • There is no additional information for this episode.

Credits

Music by Jeremy Voltz, www.jeremyvoltzmusic.com

  • CP1 (Welcome)
  • Airplane Seatbelt (Interlude 1)
  • Cowbell (Interlude 2)
  • South Wing (Outro)

Transcript

(The transcript is auto-generated and was slightly edited for clarity)

Roland: Hey J-M, welcome to another episode! How are you doing today?

J-M: I am doing fine. You know, I’ve actually had a wonderful weekend away, a chance to reflect and look at nature. I forget that often, when we’re in the city. You know, it’s full of these concrete buildings and it feels kind of constricting but just a few minutes outside of town is nature. Beautiful and bountiful, ready to enjoy, and right now it’s covered in a blanket of white and I love it. I looked over a frozen lake. I woke up to the sounds of nothing, which is a change, and I’m ready and excited to record today’s episode of What’s Your Baseline. How are you doing, my friend?

Roland: I’m doing well, and as always, I have a special treat for you which is the weather forecast from Prague in the Czech Rep awesome. We’re going to talk about that. A co-host of his own podcast. So, hello Jacub. Welcome to the show!

Jakub: Hi Roland, hi J-M, and let me just say that I really loved your intro, which sounded like in a fairy tale, you know, describing how much snow you have and I would love how it looks out there. Let me assure you that in Prague I don’t think I’ve seen any for the past two months. So I do envy you a bit.

J-M: Oh no, well, you got to get the light as it comes, right? You know, if your sun is only up for a few hours a day and that’s the time you’ve got to go. That’s why I feel like working from 9 am to 5 pm is such a weird phenomenon – it means you’re giving away the most beautiful parts of the day to your profession. Whereas, I mean, I’m happy to work at night – I’d rather work at night. There’s not as much to see.

Roland: I’m with you on that. But on the other hand (and we’re going to talk about that), Jakub, your co-host, made a better decision moving to the United States because we have clear blue sky even though it’s freezing cold outside. So you missed out, man.

Jakub: Yeah I think he brought the bad weather with him though because he just texted me the other week with his weather forecast, and in Austin where he moved to was like this crazy storm and snow everywhere which is very unlike Austin. So I think the European vibes are with him.

Roland: Well, it will change for sure. But hey, Jakub, let’s get started with the interview. Maybe you give our listeners a little intro – who you are and what you’re doing and what brought you to our show here.

Jakub: Well, I think Roland, you brought me to your show and I just wanted to say thank you very much. I really appreciate it. However, not to digress too much, I am in process mining. I am a team lead of a company, “Processand”. I work from Prague, as you mentioned, where I lead a little team of data scientists and on the side I’m also recording a podcast dedicated to process mining, which I enjoy very, very much with my co-host Patrick Bogner. So that’s a little bit about myself. I’ve been in the field for the last four years. So I’m still pretty much a newbie who is just learning his ways around and trying to learn from guys like you.

J-M: Well, we’re excited to both have you as a conversation partner and also to learn from some of your experiences. So tell me a little bit more about your past: what got you to Processand, how did you get to this story of process mining as a core part of your offering; as a professional endeavor.

Jakub: I would say persistence, because right after I finished my studies (this is basically my first full-time job) I studied in Czech Republic. I studied robotics and cybernetics, later on smart buildings. Worked at Siemens in the building technologies department, which was pretty cool. I was literally working in a boiler room and reading meters and temperature and everything, and then suddenly I moved to Germany and I was just applying for any data-related job. And I counted it (because I’m a very organized person) I got 80 rejections before I actually landed a job in Processand, and I must say, I’ve been happy ever since.

J-M: Is that normal for Germany to have a lot of difficulty finding a position particularly like an entry level one?

Jakub: I wouldn’t say so. It is difficult for a non-German though, especially coming with a limited level or limited knowledge of German. So there, I think, was the major hiccup.

Roland: But Jacub, German is easy. You know, even little kids learn it.

Jakub: Ah, yeah, that’s what they’ve been telling me since like second grade.

J-M: Well, here’s a question for you, Roland. You’ve worked in the US, you’ve worked in Germany; you’ve had to apply for jobs in both. Tell me about your comparison experience between getting a job in the US versus getting a job in Germany.

Roland: I think it’s easier for me because I came in with the daughter company of a German firm and I was already classified as what they called “International Manager”. So the whole bureaucratic stuff was significantly easier. The firm spent a little bit more on the visa, which was an E visa (an investor visa) and not the H1/B1 that you typically see and whatnot. What’s difficult and painful still is getting all the other stuff done. You know, getting your green card and all those things which are obviously independent of the firm even though they need to sponsor you for this. But I’m happy to report that I’m past that milestone for a couple of years now, so I’m looking forward to becoming naturalized at some point in time, and then I won’t have to worry about this.

J-M: Well, that’s wonderful. Well, Jakub, I want to talk to you a little bit more about things outside of work, because I know our listeners love to hear about the guests and kind of the whole person. What are your interests as a person? We always ask people about bucket list items. What if you could have you know time off from work or be retired or achieve your dreams, what would they be? What would you want to do as a high point in your life?

Jakub: That’s a very good question, and you know, J-M. I’m in this phase of my life where I really enjoy what I’m doing on an ongoing daily basis. So including my work. But I also have a personal project that I work on during the weekends. It revolves around personal finance, so you know how it is, you know, budgeting, investing, personal development and all of that. And I’ve created this very interesting platform in the Czech Republic and it’s getting a lot of traction. It’s ranking in the best business podcasts and everything. It’s like my little baby that I’ve been growing on the side of doing this day job and it’s really awesome. I’m actually in the process of writing a book, and by the time this episode airs, I will know whether it’s going to be published or not. Because there are some interesting developments over that. So cool stuff. Cool Stuff. So I guess my top item on the bucket list will really be finding the finalizing the book and having it out in the world.

Roland: Well, the interesting question is: will it be in English or in Czech?

Jakub: In Czech, I’m afraid.

J-M: Ah, well, you know if it gets big enough, they’ll translate it into 100 languages and you’ll find it all over the world and you’ll have to go on a big author tour and you can be signing all your books and will be it riding your coattails as ‘hey we once knew this guy’.

Jakub: Yeah I wish. I mean, the competition in the US market or in English-speaking markets is just through the roof. So I’m just gonna stick to my little yard here in the Czech Republic where there are not even as many competitors as elsewhere.

Roland: Yeah, that makes sense, that makes sense. But hey, let’s talk a little bit about the firm that you’re working for. You mentioned that you’re working for a company called Processand. So who or what is Processand?

Jakub: So, Processand is a German consulting company founded in Munich by what used to be Celonis employees. You probably heard of Celnois, which is a process mining vendor. One of the top on the market. The idea behind founding our firm was that ‘hey guys there is so much business going on. Nobody’s really focused on consulting services, on implementing the software.’So they just really started their own company, then soon thereafter another ex-employee from Celonis joined so there were three people. And they started getting some interesting customers because back then the base was pretty small and they knew the bosses because back in 2014/2015/2016, Celonis wasn’t the juggernaut that we know today. It was only a couple of hundred people – probably even less, back then. And they really started rolling out these consulting services. And soon after, I joined as well, as the second full time employee and I’ve been enjoying this ride ever since.

J-M: So tell me a little bit more about what kind of services that Processand offers, and from your personal perspective (because this is what we’re going to be really getting into in the conversation today), why is it important that you pair services alongside the technology implementation of process mining?

Jakub: So I already mentioned process mining a couple of times, so our core business is still in process mining and helping organizations to essentially set out on this journey. Help them with the first baby steps when they have no process mining to get them to the point where they already have the tool and they can start harvesting the insights from it. So that’s still our base. We also do the BI consulting because it’s very close once you start playing with the processes to build up a BI tool or BI/KPI measurements and so on. Also, on the side, we have a data science product and we are slowly rolling it out where we really want to serve the data science community with management of the knowledge of all the queries and everything that they have, so that’s. Essentially what we do, but our core business as Processand, we still do the process mining consulting.

J-M: So then, as a follow-up to that, I want to get into the specifics here for our listeners who could come from a variety of different backgrounds. Why do you need the services alongside the technology? Why can’t most companies you talk to do it in-house? So what’s your pitch?

Jakub: Well, that’s a good question. So luckily I didn’t do the pitches very much. 

J-M: But you believe in some of the philosophies that lead to the sales approach, correct?

Jakub: I did the pitch. I would probably mention that just because you know your processes or how they work or anything doesn’t mean that you know how to implement them. Basically, it’s categorized into these two buckets. One is the domain knowledge of the process, so you need to know your system, your ERP system. We are mostly focused on things like SAP implementations, processes such as purchasing and accounts payable and so on. So you’ve gotta know what’s behind that. But it’s not enough, because you also have to have the knowledge of what you are actually doing in the background – in the so-called data science job, where you are developing the queries and basically feeding these data models with data. And that’s where the uniqueness of the position of our company as well comes in, because we know the capabilities of the tool and what you can do with it, and we can speed up the process of implementation quite significantly. And we also have domain knowledge because we’ve implemented over the years so many processes that we know a lot of them pretty much thoroughly. And we can just jump in and say ‘okay guys if you are implementing Purchase to Pay, you’ll probably spend half-a-year before you get even decent results’. We can do this in three weeks.

Roland: That’s great. But now, obviously my question is how large is your firm? And you mentioned in the precall that you’re expanding to the United States, which is obviously a larger market. But if you could talk a little bit about that, that would be great.

Jakub: Yeah, of course. So when I started we were basically five people and the business was tough because we always had to position ourselves as this ‘cool consulting company’ and you can imagine that for some of the companies, there was a ‘lack of trust’. Let’s put it that way. However, I must say that we somehow managed. I don’t know how, looking back, but we grew into what we are now – 41 people (I counted it today) with like 16 nationalities in the company, which is amazing. So a very cool environment to work in. And we mostly still operate on the European market with our headquarters in Munich. I am now actually in charge of the Czech market or Czech office where we are growing our team. Where we are looking in the future is definitely the US and my colleague and co-host of our show on process mining Patrick is actually leading the charge in Austin, Texas, where as of now we also already have a team of data scientists and looking to grow because from a process mining perspective, it seems that the US market is is just about to start.

J-M: That sounds good. So when you talk about process mining services (and I think it’s interesting when you’re a small organization like you mentioned in trying to grow). I suspect you’re going to have to focus a little bit more on domain-specific or technology-specific knowledge because that’s an easy thing to dig into. A while ago on our show, we had Zach Bennett from VEA who talked a little bit about technology specialization as one of the great entry points. But as you’re growing, what are you thinking about? What are you seeing as a division between the kind of services you offer between technical services (so the support to platforms and the technology implementations) vs. strategic and consulting services.(so how do I get value from it and where does it fit into my organization’s strategic goals? How should I evangelize?) These sort of, you know, more traditional management consulting techniques. What do you see the division of labor being and in what clients are asking for and what you are trying to build?

Jakub: From that limited experience I have (which is about 4 years now) I see the big shift from being this platform expert (being the expert on process mining) shifting to this execution ecosystem. So basically to do the, as you mentioned, the changes, the operational management and so on which is interesting for us. Also as a company, because we have the knowledge of the tool; we have the knowledge of the domain, but we weren’t really working as a consultant. So this is something that we are also now working on because companies want to make some actionable changes based on the findings that they get from the process mining. And it’s interesting to see us developing with them –  with the companies that we work with and with our customers. We’re slowly shifting to this perspective of being a proper consulting company that can not only implement the software and help them with some basic value creation but also focusing on the bigger picture on the business process management as a whole and and maybe also on the change management and so on.

Roland: Yeah, which is obviously very interesting for us with the scope of our podcast here. But speaking of which, the companies that you work with. Why are your clients pushing for this new technology; this new approach of process mining? And how does it fit into their technical ecosystem? Is it mainstream, is it emerging or how would you classify process mining in the markets that you serve?

Jakub: Roland I really loved when you called process mining being the hipster child of business process management – I’m really gonna quote you on that a lot. So thank you for that, and if anybody asks, it’s my idea. Having this limited knowledge, I can’t really answer the full question because I haven’t been around business process management as much. When I started the job, I didn’t even know that there was a business process management area as a knowledge domain or anything. I had a hard time grasping process mining as it was. It seems to me that companies are jumping on a train of process mining a bit from peer pressure because a lot of companies are doing it. You can hear these interesting stories about how company A saved ‘x million dollars’ by just optimizing their processes and automating their issues away. And it’s also this badge of honor when you do something that’s still considered to be revolutionary and cool, and process mining definitely checks this box still. So companies really want to jump on this and actually see whether they can leverage this technology for their own use as well.

J-M: So it’s like investigative right now in some capacity. They’re kind of getting going with the hot stuff, but they’re still feeling out exactly what it fits into.

Roland: Yeah, J-M. I just spoke at a conference last week and I gave a talk, as you know, about process mining and how it fits into the bigger picture. And within the room there were about forty/fifty people, and I started the conversation with ‘hey let’s see a little show of hands, who of you are doing process mining’ and zero hands went up. And then I said, ‘okay, how many of you are planning to do process mining’ and about a quarter of the hands went up. So I think it’s still pretty new, even though obviously for those of us who work with this on a daily basis, it’s kind of normal, kind of boring, but I still think it’s very new.

J-M: And Jakub, from the experience you’ve had, that leads me to the question of resistance. What resistance are you getting or hearing or seeing – like what are ‘blockers’ in projects you’re working on saying to their colleagues or even to your face.

Jakub: So we are lucky in the sense that when we enter the discussion with the customer, they are already bought into the idea and want to do process mining. That’s because we are not the vendor. We are not trying to persuade them that process mining is the thing that they should be doing. This is usually handled by the vendor that we work for. And we enter the discussion when they’ve already said ‘okay, let’s do this.’ That doesn’t mean that it’s seamless and that we just go there and everybody’s super bought into it. There are a couple of layers that are important for the foundation of a successful project and it comes all the way from the top to the bottom. So, at least from what I’ve seen, having the management and the C-level on board as well as whomever you might imagine being involved in such projects is usually easy because those are the innovative people that want to push some changes that want to improve the organization. The problems sometimes come in the middle management that is leading the project and you have to get them on your side. These are the people that you talk most with because they are responsible for the execution of the project and you are in an ongoing discussion about what you’ve been doing, how you evaluate the success and everything. But it doesn’t end there because obviously they also have their subordinates and those are the people that eventually are going to use the tool. And getting them excited about it, basically establishing that this is not just another BI tool, but it actually possesses so much more than just reporting and KPIs, it’s really tough. And I’m not going to say that every time you’re successful with that because some companies are more welcoming of the change than others – it depends. You probably know that from your own experience that sometimes you just think ‘oh wow this was so easy. Everybody was so cooperative.’

J-M: Ah, it’s less and less though.

Jakub: I’ve had both types of customers. I had this project recently when I think it was the first time in 4 years when everything went according to plan and we delivered on time, like on the day that we said that it’s going to be done, it was done.

J-M: Well, that’s the other question about expectations because I feel like there’s some interesting conversations happening here about people who’ve bought into it but when they bought into it, what did they buy? What was the promise that they bought? And as a result, how are you having to deliver things?  So tell me about that and who’s using these insights? What are they doing with them? What value are they achieving from them? So that when they check back and say ‘yeah, you did your job’. Well, what was the job we did?

Jakub: Yeah, for us usually the job is ‘here’s your tool, use it now’. NO. There are a couple of reasons why companies are doing it. Some of the more major companies, they just want to see their processes and I’ve been involved in projects when we integrated dozens of ERP systems and it was for the first time when the company had actually centralized their view on the processes. Not only from a KPI perspective and some measurements and so on but also from the process perspective. And this is just amazing because then suddenly they see ‘okay so this is how our process looks like’. So I guess the process discovery is big. And a lot of companies are in this phase when they just want to see their process for the first time so it’s pretty interesting. Then obviously, other things such as conformance because you have different divisions and different departments that are doing process mining. So, for instance, we also have controlling teams that actually want to make sure that what their other employees are doing is correct. And the last point that organizations are interested in doing is actually the enhancements and the improvements. So they kind of know their processes, they know their pain points. They know what they want to improve and what they want to focus on. But they need more guidance. They want to be absolutely certain that this is what they need and this is how they can actually go about solving their problem. And this is where process mining is very strong and very interesting because they can actually look at a very small sub-part of the process and start really digging into that.

Roland: So does that mean, from your perspective, that the project has ended once you’ve set up the tool? Or are you also involved in the changes / the implementation that comes afterwards? So think about the role that an SI typically would play, or is your value prop ended once they have the tool and they look at it and say ‘oh yeah, we found something’.

Jakub: To stay competitive, we need to look at value creation as well. So we also shift our projects a bit in a way that we don’t just say ‘let’s implement everything’ because, youknow, we got burned doing that sometime in the past. So we’d rather say ‘okay, let’s rather focus on these 2 or 3 use cases that you say are your highest priority’ and let’s leave some time at the end of the project. Let’s put some numbers on them. Let’s put some savings potential for them and then let’s think about how we can solve the problem. And then you open up different paths that you could take. And when I was thinking about it, I thought that there are basically three paths. One is the people path, so basically influencing what people are doing and the way that they are doing it. You have the clear data. You can just go to your subordinate and tell them ‘look, this is not the way you should be doing it’ or call your vendor, call your customer and just improve it. The second thing would be the process path. This is already somehow system related or IT related when you basically make changes in the process. We are not specialized in doing that yet. We have some knowledge of things, we can pinpoint the problem in the master data or so but the customer needs to do them themselves. We don’t really support them in these problems. Then the third part which we can already use is the IT path. And that’s basically when our knowledge of the tool (and we mostly work with Celonis as a process mining vendor), that’s when our IT knowledge really comes in handy because we know what the tool is capable of and it’s really invested a lot in automations / alertings. But also there is this integrated machine learning Workbench, and I can already tell that at least when we are using it it’s not just a buzzword, and that we have some interesting use cases where we actually used an actual machine learning thing to produce some results.

Roland: Well, just to have it said, because we’re obviously an equal opportunity podcast here, there are obviously other process mining vendors out there who are on par with those features. And obviously you, dear listeners, have a choice in choosing your tools. But to come back, we mentioned it earlier, you’re running your own podcast called Mining Your Business (shameless pitch guys listen to them). Patrick and Jakub do a very good job here. But one thing that I noticed (I’ve followed you for quite a while now and listened to the episodes) is that you had that epiphany that there’s more than process mining that there’s that whole discipline around BPM and I enjoy those episodes when you bring in old farts like us. But the one thing that I noticed is that there’s also something like a renaissance of process management for young people, which I think is super positive. Jakub, do you see that too or what is your perspective on this?

Jakub: Yeah, hundred percent. So what I think young people really want, especially when they are looking for their first job or first career opportunities, is to make an actual difference. The data and any job related to data was like the hottest job over the last couple of years. If you look up data scientist, data analyst or any hard skills that you use for manipulation of data, be it Python, SQL, R, you would see a huge spike in positions like that. And the great thing about working in process mining and especially being a consultant such as our company, is that you can actually take data and, you know, do essentially whatever you want with them because you have access to them. You have the opportunity to pinpoint problems and also you have the tools in your hands to actually fix them. And I’ll give you one example. So I mentioned that we use this machine learning workbench for one of our use cases when we are basically checking duplicate payments. And because we can consolidate the data from different ERP systems into one environment which for most companies was unthinkable just a couple of years ago, we can run our checks on a lot of different data from a lot of different systems. And I just had this demo a week ago when one of our customers wanted to see that so we implemented it and just during the demonstration, we found a duplicate invoice that wasn’t called by their system because there were some just different characters that the the system didn’t catch. And we found and stopped an actual saving of 60,000 Euros, which for me just at that moment was pretty cool. And I guess young people really look for opportunities like this so that they can leverage the knowledge that they were taught at school and go into the real world and make a difference.

J-M: And I actually have a story about that, specifically about young people making the difference. I have a really good friend, a very close friend, a wonderful human being who has a PhD in engineering (biomedical engineering) and he found that something for both employability and for drive to make a big difference, as much as doing biomedical research was actually turning that into data science. So he went back and got a certificate in data science and he became a fellow at a data science institute as a career move as a you know 30 year old man. And that that was a huge, positive shift for his career and has landed him where he wants to be. And it’s funny that was where he saw the opportunity to do the most good. To use data and to have that as his scalpel to make a real difference to the business space. And I think that’s really cool taking that sort of approach. And the question is where are those people getting that information? Because I feel like right now, Data Science does exist in some places as a degree but I feel like we talked with a few of our different guests. Wow, a lot of these sort of skills come from on the job. Can you tell me about what you’re seeing and particularly when you’re looking at hiring up a team as you are in the Czech Republic, where are you seeing those skills come from? Are you looking for people who’ve had to do it in industry or are you now seeing this explosion of programs that promise the skills that you need.

Jakub: I love this question because yes, I am involved in the hiring process. And we did a little bit of A/B testing, checking what’s going to rank best, how many candidates each job listing is going to give you. I can tell you that when you write ‘process mining architect’, you’re going to get zero replies. Because nobody still knows what process mining is, especially when you get out of our little community. Unless you work in the field, process mining is still very much unknown. What is not so unknown is the position of a ‘data scientist’ or ‘data analyst’, but behind these things a lot of different things can hide. So then it really comes to the point where you really need to explain what’s behind that, and maybe coming back to the question, I also noticed when looking up the new candidates that what universities offer is to give students a little bit of insight into everything. So they get these hard skills and they get the first experience with working with larger data sets and that’s essentially all we need. Because no school (or at least no no school that I know of) will teach you how an ERP system works or god knows how SAP works – not even I know that after 4 years of working. So this is something that we basically teach them, but we need our data scientists to at least possess this general, I would say IT affinity so that they are able to work with the data. They know what the data structures are and can write some basic coding to work with it.

Roland: Yeah, that makes sense, Jakub, and I think it’s also a good thing because it means it’s an opportunity for growing forward. When you have people who learn about these tools, about these techniques, about these approaches, and then think ‘hey I can implement that in my organization’ which is obviously good for everyone who’s involved in the currently smaller ecosystem of this, and the pie will just grow. But to give you some time to reflect on this, our dear listeners, we’re going to play a little bit of music as always after a segment but before we go I like to think about how process mining might fit into your organization. And how you can use process mining insights to improve your current situations. We will have wonderful music from Jeremy Voltz again and we’ll see each other in a few seconds.

Musical Interlude: “Airplane Seatbelt”, Jeremy Voltz

J-M: All right folks. Thanks so much for listening to this wonderful music and taking a second to think about how you and process mining might become best friends. But talking about good friends and making good friends, let’s go back to Jakub. Because Jakub, I know that you work on a lot of process mining projects and I’d love to get sort of two layers of discussion here. Number one, I’d love to know: how do those work? Go through the different phases of a project that process mining would sort of be the centerpiece of. And the second layer is: where does this fit into an organization and how do these projects get slotted into your typical project delivery lifecycle? Are they part of large-scale initiatives or are they just one-offs. So let’s start with the breakdown and then talk about context.

Jakub I like to say that a lot of companies are expecting process mining to be this ‘out of the box’ solution that you just basically plug and play some transformations, extract the data within a couple of hours and you just do process mining and save $10M. It’s not like that.

J-M: It’s a self-driving car, right? Process mining? That’s how it works?

Jakub: Yeah, you just don’t need to know anything.

Roland: Just to tell a little side story, I had a client who wanted to have a look at the procure-to-pay / order-to-cash process and was using Oracle. And I said ‘hey, that’s great. You just had the implementation, so you know what the process is.’ And they said no, we don’t know how it’s customized. We’re looking at the vendor to help us with figuring that out, as if we would have obviously Oracle experts piled up like crazy.

J-M: Yeah, we have experts in your business. Like we know how YOUR Oracle works? No, we don’t, and you don’t either? That feels a little rough. So it’s not plug and play. You’re saying that you have to bring people together with technology. So tell me how that breaks down into phases of a project and where do you introduce different components.

Jakub: For us, when we get the process mining project and we have a brand new customer who wants to do process mining for us, it always starts with what we call a kickoff. This is the moment where we generally find out what the customer wants. It’s important because we don’t know what kind of processes they want to look into. We don’t know what kind of database we are talking about or what kind of source of the data we are talking about. If there are some external systems and so on. The security layers etc. There are a lot of things that you need to ask at the beginning. And from kickoff to actually starting with the implementation, many months can pass, because you can imagine that some large organizations won’t just let you extract their financial data from their system and just do process mining on them. You have to go through a lot of approvals and security checks. I’ve been on a project where it took six months from the moment that they bought the license and wanted to do the process mining until we were actually able to connect to their system because the pipeline from the minute that you knew that you needed the data until you actually were able to acquire it was insane.

J-M: And they’ll say ‘well, why is it taking so long for you to run things?’ And it’s like ‘well listen, you put these protections in place, and with good reason, but as a result, when we need to pull this information that you’ve protected so carefully, you’re going to have to pay a little bit in time to comply with your own regulations. And you don’t want to try and go around those because now you’re implying that the process isn’t important.’ So now you’ve gone from kickoff, to the first moment of the project tell me a little bit about how that starts when you’ve got the data set. *Chef’s kiss*

Jakub: So we learned that we should postpone every other interaction with the company until the moment we have the data. Because we don’t want to learn what the users want to mine and then wait six months before we can actually do anything. I can tell you right now that you lose all the attention. So you don’t want to do that, right? So you want to make sure that you have the data. And when you do, you need to go talk to the business. So as I said, process mining is not anymore just about let’s show purchase to pay and we are done. We need to focus on what the customer wants, and luckily they are more and more educated so they kind of know what they’re jumping into. They saw all of these value creation workshops, they know what other customers are doing, so they kind of have an idea. For us, it’s important to listen and to ask them: ‘why is it that they want to do this specific thing?’ ‘What would they consider to be a success, and how do they want to measure it?’ Get all the necessary information for us to build them a reporting tool or reporting system so that they can get the information from that. So we really have this  scoping workshop while we get all the definitions and then we just go code. This is the moment where actually most of the heavy lifting happens. This is where we have the data. We build our transformation. We build our activities in the process. If we have some external system like something scanning for invoices and so on we build it in and we usually take 2 to 3 weeks before we go back to the customer and we do this handover workshop. We show them their process and we walk them through it, let them ask all the questions, and like give the tool to them. However, it cannot stop there because you know when you see process mining for the first time you have no idea what you’re doing. And I love this because I usually show them how you can crop the process and look into different variations and they’re like ‘ah okay, could you do this again? Where do I click?’ So this is usually when the fun part happens and you cannot really leave them hanging from there on. 

Roland: So I have two questions on that, Jakub, because I’ve seen this in the past. The first one is the question where ‘I don’t believe you about what I’m seeing here. I think you made it up. It’s not my data.’ How do you counter these types of questions?

Jakub: I say: ‘prove me wrong’. If they say that it’s not true – let’s just pick a random example. Let’s pick one invoice and let’s take a look at that into their system. They’re saying ‘it’s not possible that it has been open for six months.’ I say ‘okay, let’s take a look. Here’s the number, take it, go to SAP and take a look at that.’ And they’re like ‘oh my god, you are really right! It has been open for six months. How’s it possible?’ Well, I don’t know, you tell me! 

Roland: Yeah, fair enough. But I think it’s a point worth mentioning. To ensure a successful project, you have to make sure that you convince your client that your data is right because otherwise I think you just fail from the beginning. But the second part of my question is: when you go in there (assuming you have a very immature client), do you have some standard analysis types that you bring to the table – things that you always do at the beginning of a project? Things you always look for? And if so, would you mind sharing those?

Jakub: Yeah, absolutely So, one of the easiest ones is the changes. So when you are looking into any process, usually the good practice is to have it automated as automated as possible. And seeing the manual touches in the process is usually one thing that everybody begins with. Because you can identify these changes. You can say how long it took, what kind of impact it had, how many cases / how many items in your process were affected by this change and very easily put a number on that. So you can say ‘okay, if somebody takes 5 minutes for each of these changes and you have 100,000 of them, how much time would you actually save and eventually how much money would you save if you didn’t have this change if you just reduced it by 50%’ So that’s one. Automation is another big use case. You know you want to show them how well automated their process is and what kind of impact it would have, if it were more automated. And then cycle times. So how long does it take for your case to get from point A to point B. And if it’s too long, why? For what vendor? Where does it take so long and why is this the case? So these are, I would say, three major ones that easily demonstrate the power of process mining.

J-M: Yeah, here’s a question for you about time, because one of the things I haven’t heard you say is cost. I know time turns into cost specifically. But how do you quantify the dollars, or do you actually get the customer to quantify the dollars from their business domain expertise and knowledge?

Jakub: So we do this in the last phase. After we give the data to them and hand over the dashboards, we go into this feedback loop with them where we try to get feedback on everything. We usually take a couple of weeks. And at the end of this process, we do the value creation workshop. And that’s exactly where we are trying to put actual numbers and actual dollars on the findings that we’ve made between the handover and until we meet again for the value creation. I generally don’t like putting numbers or putting the dollars on anything because it’s just so easy and very often so wrong to just say ‘okay if I save this, it will save me $5 per case and times 100,000 cases, we’ll save half a million.’ And everybody just gets excited and says ‘yeah, let’s save half a million’ and then you find out it’s actually not a problem and you know, we can’t save half a million because it actually costs us only ¢20.

Roland: Yeah, and there’s obviously a lot of things that go with it. When you think about cost savings, you could say ‘oh look, you’re having three different systems here. Wouldn’t it be great if you could retire two of the three?’ And you have no clue if those other two systems are mission-critical in a completely different process and you’re actually just talking nonsense. What I was thinking, Jakub, when I was listening to what you were just saying, was what about if I had to write a statement of work? I understand we have the 4 phases – the kickoff, getting the data, doing the replay to the clients, and then the value creation workshop. But how many iterations do you have in your project approaches? Because from my perspective, the clients I’ve worked with (and maybe I had the wrong ones, I don’t know), they were always asking for more. ‘Oh can we do this?’ How do you get a handle on this so that you don’t have a project (even though it might be interesting to see where they’re going) that gets the scope creep to a so-called ‘nirvana’?

Jakub: Yeah, I guess it’s about hitting the sweet spot. When the customer is left at the end of the project and they are just craving for more because suddenly they see that they could do so much with this, if only they knew how to do it themselves. Then they think ‘I wonder what company could help us? What about the guys that actually implemented it?’ 

Roland: Yeah, that’s a very interesting question. 

Jakub: I’m happy to say that a lot of our customers stay with us long-term. So for us, it’s not usually one and done and we do like to support them not only on the processes we’ve built, you know, already going more into the depth into some use cases. Because I love to say that process mining is never done because you could always implement something more, enhance something more, catch some bugs, look into the bigger details, and so on. But you can also start with new processes because there’s just so many business processes that you could be doing.

Roland: And it’s great to hear that you’re building up the longer-term relationship with your clients because I think that is one of the true success factors in consulting if you have that relationship with your clients. But speaking of clients, what should clients expect when working with you, and maybe more importantly, what should they prepare before you put a foot into their building and start working with them?

Jakub: Well, I would say, get ready for having to work. Because we will bring a lot of work for you. It’s not that we just do the implementation, pinpoint the problems and help them with value creation. As long as the customer doesn’t act on the findings, nothing gets done. And they need to establish a Center of Excellence, appoint some people who would be responsible for looking into the use cases and getting the value out of them. Because if we are missing this dynamo, this person who’s driving the initiative, it’s very likely to fall or to be forgotten at some point. So I would just say, if you’re going to do process mining, be prepared that it’s going to be a long term commitment.

J-M: And that leads me to a follow-on question (in general though) because I think that you’ve had a chance in just even just the 4 years you’ve been working on this to see a bunch of different client circumstances. And also to see technology evolve and develop. We want to get some lessons learned and insight for people who are looking to get into this or even companies who are struggling to figure out whether or not they should talk about process mining. Whether they should get into projects. What are your lessons learned from entering this marketplace and what are your most successful and challenging engagements? What are the things you see that are common about them? Help prepare companies, prepare people for their expectations with some of your best tips and tricks.

Jakub: Yeah, that’s also a good question. I would say if you are interested in lessons learned, just go listen to our podcast: Mining Your Business. We have a whole episode on that. There are a couple of them, and the longer I work in the field, the more I realize this first one, and that’s actually listening to what the customer wants. For me as an engineer, it’s very easy to just forget it or just do my thing and just leave off because I know how it should be done. But when you are starting to feel more like a consultant, you have to understand what the customer wants, and just because you’re gonna build something that you think is amazing / is cool, doesn’t mean that it’s actually what the customer wants. So understanding their needs and trying to help fix their problems, whatever the problem might be.

J-M: And so if you’re looking to get into this career. Let’s say you’re a student. Maybe you’re in high school or even early in University. What sorts of experiences should you seek out? And thinking about somebody very early in their career, what kind of jobs should you get that will lead you down this path? Jobs that would let you join something like Processand and you in this career path?

Jakub: So, as I said, in our hiring, we don’t necessarily look for experts. We actually like people who are just fresh out of school who have some experience with the data. People who worked on a diploma thesis or something that’s somehow data related where they had to process large amounts of data and had some output out of that. So we do like these candidates. So if you want to work as a data scientist, I would say start with that. Try out different projects. There are so many online courses that focus only on data processing and data analytics – do some of these. See whether it’s something that you like and yeah, then just apply.

Roland: And once you’re hired and obviously you’ve done some projects, where do you see the growth path for those candidates? Because obviously you’re working in a consulting organization. It’s not academia anymore. So what would be a perspective that somebody would have who starts today at Jakub’s team at Processand.

Jakub: I would say that you can essentially take at least two paths that come to mind. You can become a domain expert – when you really know the backbone of the ERP system, be it SAP, be it Salesforce, be it Oracle. Something that you do the data mining on. That’s one thing, and that inevitably comes should you work on these projects because you need to understand how the relationship works and so on. But then you start to develop this feeling for the processes. So after you build a couple of processes, you know what kind of problems customers have. And this is the fun part because it just opens you up to so many new things that you could be doing because you start having answers to these questions. You know the tool. You know the domain. And you can develop very interesting technical things to actually address these issues. And as I said, we’ve started using some pretty advanced machine learning algorithms to answer some of the questions that our customers have, and there are still so many things that we could be doing with that. So one thing is definitely becoming an IT expert. And then there’s the second path, which I personally took and which I really enjoyed. That’s more of the consulting thing, where you talk to the people, you lead the projects and you are that ‘people person’ when you are helping your IT team to grow and to be able to translate what the business wants. But you are also talking to all these business people and learning about their issues and trying to find the solution that could work for them.

Roland: Yeah, and I’ve noticed that not only in your podcast, but obviously in others and obviously in our daily work. Once you’ve been into that process field for quite some time, it’s not the technology; it’s the people who are the ‘problem’. And I find it so funny listening to other podcasts who say ‘oh, we’re the technical podcast’ and after 5 minutes they start talking about people and their interactions with them. Because at the end of the day, it doesn’t matter how much wallpaper you’ve made, how many process models you put on the wall, and how many black boxes you have in the basement. If the person behind the screen doesn’t want to do the work, they won’t.

J-M: Yeah, well that leads us to the end of this segment and Jakub, it’s been incredible so far. We’re going to have a little bit of a conclusion after this, but we wanted to leave people with a little bit of music and a question or a couple of questions to take you away in your thought time. So first and foremost, think about what you’re doing right now. Whether or not you’re a student, early in your career, mid or late in your career, what are you working on right now? And how does that touch on data / large data sets and the extrapolation to process? Could mining be for you, and if it is, when do you think you’re ready to start that conversation as a topic for your thesis for your project for your career or as part of a lifestyle? We’ll leave you for a few moments to think about these things and we’ll come back with our conclusions and the end of the episode.

Musical Interlude: “Cowbell”, Jeremy Voltz

Roland: Welcome back. Very nice conversation with you Jakub, and it was great talking about your experience, your firm that you’re working in and the interesting things that you do and obviously all the benefits and approaches that you bring to process mining and process analysis using that modern technology. But the question that is obviously on everybody’s mind is now that they’ve heard you, how could people reach you? 

Jakub: The usual suspects. I’m on Linkedin and I assume that I will be tagged under the What’s Your Baseline beautiful website with all the information about what I do in my free time and how I can dunk the basketball and everything. I’m on Linkedin (Jakub Dvorak) and we have our podcast and you can find it on LinkedIn as well. It’s called Mining Your Business. Or we have our website, https://www.miningyourbusinesspodcast.com/. If you’re interested in what we are doing as a company, we have a corporate website, http://Processand.com so you can find us there and leave us a message.

J-M: Well, thank you so much, Jakub. This has been a really great conversation, and a great delve into process mining and the consulting that makes it happen. Now, speaking of making it happen, all of our wonderful listeners make it happen every week by liking, subscribing, sharing and joining us in these amazing conversations. And of course, you can find out lots more about this podcast, and a ton more at https://whatsyourbaseline.com or for this episode specifically (for the transcript and show notes) at https://whatsyourbaseline.com/episode21. Well, folks, until next time, I’ve been J-M Erlendson.

Jakub: And I’ve been Jakub Dvorak.

Roland: And I’m Roland Woldt.

J-M: And we’ll see you in the next one.