#19 The Process Mining Q&A Episode Part 2

On the order or on the customer record, so we can take this even further. So when problems occur, we can take immediate action without necessarily the human in the loop and can avoid even more problems or can react to these problems even quicker. Yeah, so yeah, perfect to combine it. Yeah, so the next one we received here on our list is, you know, what kind of problems and solutions can we discover with process mining? And yeah, Yeah, so we covered that, I guess, in our previous podcast a bit, use cases for process mining. So which one do you think is a good one to start with, Arno?

I mean, there's so many examples, but it could be if you look at, let's say, business development, that's a good one. I think business development is quite an important discipline within your business to get right. And perhaps, you know, there's areas where the and within your business development or even with your accounts payable where, you know, having problems with a specific supplier or problems with a specific account manager, for example. So process mining allows you to pinpoint who is causing these problems and also how big these are in terms of business impact. You know, these types of problems. And you could use these tools built into the process mining tools to be able to look at service level agreements, make sure that those are adhered to. And you'll also be able to spot the supplier or this account manager that's causing your organization sort of significant problems. Or there's errors, or perhaps there's potential fraud. With inside that realm, that's a really good early warning system, you know, for your business. You also might find that, you know, employees in specific geographic areas have not really understood how the ERP system works properly. And, you know, that might cause a lot of problems for you, which, you know, adds up to a very large business impact. And, you know, that might then identify the need for more training. For those employees within that business area or unit and how to use that ERP system. And of course, that would result in significant cost savings to your business. But if you didn't have process mining, how could you identify that training is needed for that specific location? You know what I mean? Unless you actually either just stumbled incidentally across it or it sort of got caught by your, your training program, you know. So with this, you can be very tailored and very precise.

Yeah, absolutely. And then you can really uncover as well, even, um, yeah, if someone is, uh, yes, not familiar with the system and then the potential delay it has for certain process steps. So then, uh, yeah, I can really come up with all sorts of information to measure the impact. Yeah, and another problem when employees not necessarily follow the main process. So usually we all have this ideal process in mind and lots of department heads, they're sitting down and designing all this perfect process in Visio and then doing the introduction to this process to the department. But there are so many so many variations sometimes. And yeah, because everyone is sort of living that process slightly different, sometimes it's living it different. So yeah, this is very important to really find out where all these variations of processes are. So these variations are causing issues generally. So, so, but with Process Mining, you can then uncover these variations and say, okay, how can we tame it? And then if you're thinking about robotic process automation, So we could then say, okay, how could we potentially get the process again sort of more unified again, more streamlined again by maybe just passing it on to a bot because people are getting maybe confused anyway, they don't like the process maybe. So if we pass it on to a bot, the bot can concentrate on it, will process everything. But yeah, so all those kind of variations and potential for Robotic Process Automation, we can also uncover with, with process mining, yeah, we could potentially even record lots of different sessions of users and to identify how are people working within the process. And by understanding all that stuff, we can then come up with really, really good rules when to use a bot and fill our pipeline as well for our RPA projects.

Yeah, that's true. And I think, you know, RPA in itself, you know, can be actually quite costly. To implement if you don't really have a sufficient understanding of your processes. So, you know, process mining shows those variations like you said, and also shows you, you know, what's the common ones. And the common ones are usually the ones that, that's really good for RPA. And so this is really, really good tool, a really good way to, to to improve your processes, standardize it, and kind of identify those common ones that you can actually push through your RPA bots, for example, rather than doing it manually. And if you still have to do it manually, I think that's fine. You can use process mining tools to identify these processes. You can then from, let's say, operationally understand how you need to change them, and then you can measure if people actually adopted the changes. This is one thing telling people, you know, this is the changes we need, but how do you measure that? You know, how do you know that people actually are living these processes like you say, the way they should? And process mining do give you that, I call it telemetry, it's almost like a business telemetry, that real runtime telemetry that you could measure and say, you know, we have adapted our training programs and we are seeing the data is showing us that people are actually doing the processes the way they should be. Hence why we see the bottlenecks reducing and the problems in these processes going away.

Yeah, and we can really identify as well, yeah, all these processes are running maybe in different regions or process. Yeah, we can really identify as well all the touchpoints of in our processes who's doing what and then really identify lots of these opportunities, as you said, with the common ones. And then this can save millions really if we do it in a larger scale and then the right organizations. We can reduce sort of the touch points in the processes by first identifying these variations and then put automations in. So So it runs much, much, much smoother.

Yeah, good. So another one we received is a common one where, and this is sort of reflects the reality where people say, well, our data are in different systems. Can we still use process mining software to connect these systems and combine the data from different sources to gain these insights? And the answer to that is a categorical yes, because process mining software do have connectors that enable you to transfer data between the various systems to unlock this data and enable the software to connect to these varieties of single source of truth systems. So these connectors then allows your process mining analyzer to look at data across your enterprise So it enables you to really look at data quality and also deploy real-time analysis across your business dataset. So if you do have disconnected data systems and applications, so you might have several SAP instances, which we see a lot. You might have Salesforce, you might have custom SQL databases, Excel files, So using process mining software and the connectors they provide, you can bring all of that event data from all of these platforms and bring that into your mining model. And that's really, really useful for you to see that sort of, that bigger picture. So it's not just kind of targeting one specific silo.

Yeah, and I think it's a— and I think this is the key actually. It's not that we can generally support with process mining looking into other systems. I believe this is fundamentally important to do that actually, because how would you otherwise really determine your end-to-end process, how it really runs? So if you only have the chance to look at the individual system, that's great that you know everything what you're doing in Salesforce is not working or where it's not working. I think because we have so so many different systems these days. I think when we're very modern companies, they have these 20, 30, 40 different cloud applications and where they sort of have one process end-to-end through these 20 applications. And if they wouldn't be, if process mining wouldn't be able to gather all that information from all those systems, it would just give you really not the true picture and you wouldn't really see where things are going wrong. So it's fundamental that process mining tools can have as many connectors as possible. So this is key. So, and yeah, with that you have the chance to really identify everything and it goes really that far. So I mentioned just the recording earlier on. So there's process mining, but part of the process, there are lots of tasks. So doing task mining, process mining are now sort of terms very often used together. So you really have to identify as well not just what is happening in the silo systems, really what's happening also on the desktop of the user. What kind of tasks is this user doing on a day-to-day? Maybe it is purely Excel, they're working on half of the process and then it moves into another system or a few other systems, but that activity what happens in Excel we also want to know about ideally. To get the full picture.

Absolutely.

Good. So the next one is a typical one. So we obviously talk about this topic a lot, and then customers think, yeah, that's great. So what is required from a business in order to use process mining? So what would be needed?

Well, of course, you know, an easy place to start is if you have a centralized ERP system or a single source of truth system, you can, of course, combine things using connectors. And most process mining tools obviously support that aggregation of data. But I think, you know, from a very practical perspective, that there's really sort of just 3 steps to to exploit process mining. And really, you know, the first one is connecting your data to a model. And, you know, this is where you take your process mining tool, you connect it to your data source or data sources, it then brings that data into your model. And that's then ready for analysis. So it's really kind of just sucking that into that model, process mining tool, understanding the makeup of that data, and then you're ready for your analysis. The analysis part is really, I guess, where the magic happens, so to speak, whereby you're process analysis cycle starts on your data. And, you know, this could be a longer sort of activity. And this is where, you know, your process mining analyzer would really present you with, you know, various things in your data that is of interest. So that's sort of taking that business telemetry and bringing it into life to give you insight. So once the data is into the model, the model then turns into insight. That's where you actually see the telemetry come to life. And then the third thing is really, you know, the third part of the journey is really where you start making changes to your process based on the insights you've gained from from the previous step, and that's where you unlock the value of process mining. So there might be a particular process that you focused on. You might have gained an insight that it— there is a bottleneck because of a problem. You might then decide to actually change the process and deploy RPA software. And then again, if you look at your insights for your sort of the new data that, or the new telemetry that comes through, you will see that problem go away because the bottleneck is reduced. So it's really quite simple, you know, these sort of 3 steps, again, kind of get the data source, put it in the model, extract the insights, you know, tweak the process and really see the changes and extract the value from that.

Absolutely, yeah. I think that brings us to an end of our episode. And I think, yeah, it's a really good time to take a look at your processes with process mining. So there are lots of easy ways into it. So proof of value will be a really good start. Reach out to Arno and myself via the automation guys, Don website to just get things started. If you want further information on it, we will have more episodes coming up on process mining. Yeah, it's good to also to look back at our previous process mining episodes and intelligent automation episodes so you can see how we bring all these things together for maximum value. And yeah, so that's it for today. Thank you very much for listening and Hopefully this episode was useful for you and will get you started with process mining. And as always, thank you and bye-bye. Unfortunately, that's it again with this episode of the Process and Automation Podcast. If you like this episode, please give us a 5-star rating and don't forget to subscribe to this podcast so you don't miss any upcoming episode. We hope you will tune in next time, and until then, let's automate it!