This post includes the full Transcript of Process Mining From Automation to Improvement from Ashling Partners Webinar on October 1, 2020 and the YouTube video of the entire educational webinar: Leveraging Process Mining For Process Understanding, Automation, and Continuous Improvement.
In this webinar, Rudy Kuhn, founder of Process Gold and now head of Process Mining at UiPath, will show how the interplay of Process Mining, RPA, and all the other components of the UiPath Hyperautomation platform enables the sustainable improvement of processes and maximization of efficiency. Tyler Safranek, head of Process Mining at Ashling Partners, will demonstrate the power of UiPath Process Mining by showcasing a specific common use case in ServiceNow processes and show how to connect a process model to business outcomes. Speakers include Tyler Safranek, Tower Lead – Process Mining & Discovery, Ashling Partners.
Transcript Introduction:
Marshall Sied
Awesome. So good morning, everybody. This is Marshall Seid, I am a co-founder of Ashling Partners, this is going to be a very exciting session today. We’re going to go through introductions on who’s going to be presenting today, really excited about the topic around process mining. You know, being in this Hyper Automation, Intelligent Automation and RPA space. This is one of one of the more emerging and exciting categories and capabilities that we’re seeing in business operations. I think at the end of the day, one of the largest bottlenecks for not just automation, but transformation and continuous improvement is really around process understanding, and not just doing it kind of in a in a snapshot fashion, but in a continuous fashion. So we’re go through some kind of trending topics today. So very excited to have everybody as a part of this session. So I’m going to ask everybody to do, quick introductions, with the caveat that I think Brandi Corbello is still joining the call, but she should be here by the time the panel starts. So Rudy, can you kind of just give us get us kicked off with a quick, quick background and introduction?
Rudy Kuhn
Yeah, sure. So first of all, thank you for inviting me. Rudolph Kuhn usually go by Rudy. I’m the guy who back in 2010, founded process coach in Frankfurt in Germany. That’s where I’m based right now. And our company was acquired by UiPath. In actually almost one year ago. So in two days, it will be our anniversary. So and after the acquisition of Process Mining by UiPath. I’m the Vice President for data analytics and our Global Head process mining ambassador. It’s my great pleasure, you know, to discuss today with you the role of process mining in our hyper automation platform.
Tyler Safranek
Yeah, everyone. I’m Tyler. I’m the tower lead for process mining and process discovery at Ashland partners. I’ve been doing process mining projects now for roughly four years, started at 3am doing process mining in midst of their SAP transformation. And so really learn, you know, the value of process mining, not only in terms of process transformation, but also kind of global business transformation. So I’m excited to talk about that today. I’m sharing with you my experiences.
I’m Brandi Corbello, I’m a senior director of transformation at Cushman and Wakefield, really kind of leading the way in our hyper automation journey. So excited to be here.
Marshall Seid
So Salamn Khan is in and Brandi’s organization, also leading transformation. Part of that is automation. But part of that is broader continuous improvement and transformation as well.
Agenda
So I’m going to go through the agenda and then I’m going to hand it over to Tyler. Really what we wanted to do is set context and set the stage around what it means to be process centric, giving historical examples but also talking about what infusing process mining does for an organization from an operational excellence and speed to market perspective. This is very, very visual type technologies and capabilities to Tyler’s going to take us through a quick example of going against a widely adopted ITSM enterprise application and ServiceNow.
Rudy is going to take us through the UiPath vision for process mining. So that includes process mining as a standalone capability, but how that integrates into the hyper automation platform. And then Brandi and Sal are going to take us through how Cushman and Wakefield is leveraging process mining today and their deal management process. And then we’re going to conclude with a panel. So feel free all participants are muted from a chat perspective. But feel free to post questions. We’ve got some that we received prior to the event. But we’ll try to get through as many questions in the panel as possible as well.
Becoming a Process Centric Organization
Tyler Safranek
Awesome. Thanks, Marshall. So yeah. Hi, everyone. Thanks for joining today. Appreciate your time. Today, I want to talk to you about a topic that’s near and dear to my heart process mining, of course, but you know, more broadly about how process mining can help you become a more process centric organization. Beyond just, you know, finding new automation opportunities, it really helps drive a cultural shift in terms of how you manage your processes, and manage your company overall. So I’ll talk about a quick historical view of process centric organizations define really what a process centric organization looks like, how to change it, and what it what it’ll look like after you’ve implemented that change with process mining. So, without further ado, let’s get started.
So, history really has always favored those who focus on process, starting really with, you know, Ford, and the innovation in terms of developing the assembly line in the manufacturing processes that we know today. So, their vision really was one of process that produced a specific output, it could be, you know, change slightly to produce slightly different outputs. Then we’ve got Bell Laboratories, which was really the first in terms of looking at a process and determining, you know, statistically if, if that process was in control, and they really kind of laid the foundation in terms of looking at a process. From a six sigma perspective, though it didn’t coin, Six Sigma at the time. Next, we have Toyota here, Toyota is well known for their mantra of the Toyota way, which is really a culture of continuous improvement across the entire organization. One of the fundamental mantras of took the Toyota way is that the right process will produce the right output. So making really that process to the output that you get from your, from your processes. Next, we see Motorola, obviously a huge producer of bones and various technical technology, they were the first really to coin Six Sigma and solidify the tools and the certification process that came with six sigma. And so obviously, they were very interested in looking at their processes from a statistical measurement standpoint. And that really helped them drive and become the leader they are today. GE then took the Six Sigma framework and applied it broadly across the entire business. They basically had a mandated Six Sigma training that everyone in the organization would take to ensure that they really had that continuous improvement mindset and could all operate and talk kind of the same process improvement language.
Next, we have Dell here, Dell is really known for their lean manufacturing operations, especially in their heyday, it was estimated that at their particular locations, they would have maybe only two hours of inventory on hand, and across the whole company, 72 hours of inventory. And so they wouldn’t have been able to do that without some sort of process focus and process, measurement tools there. And lastly, our poster child for process these days is really Amazon, their ability to use technology process, and people to drive really an amazing customer experience and extremely lean, and just in time operations. So all of these companies, of course, are kind of towers in their in their respective industries, and really have reached that level by using process centric focus. But of course, we’re always iterating innovating and so excited to talk to you a little bit about what those innovations will look like today.
Fundamentals of a Process-Centric Organization
So fundamentals of these Process Center organizations, first and foremost, all of these organizations fostered this continuous improvement mindset. across every level of the organization. There’s an openness and there’s a transparency that comes with that open feedback between manager and team member and vice versa feedback up the chain to in terms of ideas that can be implemented to improve those processes. So that’s, you know, that’s a cultural shift that really needs to take place for you to reach that process centric level.
Next, there’s a fundamental understanding really that output is a function of people process and technology. Sometimes we get fall into traps of thinking that output is really just a function of people or technology. But the best organizations understand that that output is the people process technology, the holistic approach, which needs to be considered in order to really ensure that your output is meeting standards in arriving on time.
Next, we obviously, process centric organizations have standardized, transparent and measurable processes. Most companies, of course, have standards for their processes. But they may lack sort of the transparency and the ability to measure their processes effectively. And so process mining, as I’ll talk about today, is really going to help you achieve those that transparency and the ability to measure your processes. And those are key components to really being able to, you know, process centric. And kind of Lastly, business processes are really, business process decisions are made using empirical data. Of course, process mapping sessions are valuable, you get the subjective opinion of your experts in the room. But you need to really combine that those subjective opinions with empirical data before you can ensure that the decisions you’re making about your business process are the right ones that are going to have the desired effect. So, you know, organizations that are process centric are always going to look to the data to ensure that they’re making the right decisions about their processes.
So why would you want to be process centric? If I hadn’t made the case already, in terms of seeing the companies that are process centric, and the success they’ve had, I’ll just run through a few points here. In terms of what I see about what are the advantages of process centric organization. First and foremost, you’re going to be able to increase your output quantity or your quality without adding headcount, you’re going to look for ways to do more with the resources that you have currently, you’re going to be able to respond to change quickly and effectively. Because you’re going to have that transparency, to see when changes are happening in your process, before your output is really affected. So you’re going to that transparency is really going to allow you to be agile and respond to change quickly. You’re going to really increase your employee engagement, because everybody’s going to have that feeling of responsibility for the output, their actions are all going to be tied to the corporate objectives. That transparency, again, is going to allow employees to understand exactly what their day to day jobs are contributing to at the high level, so that employee engagement is really going to go up and you’re going to have I think, much happier employees who feel a sense of autonomy and entitlement to the output that they produce.
And then finally, you’re really going to continue to see the new technologies that are emerging in terms of automation and continuous improvement. As you’re always looking for continuous improvement technology. Again, it’s a core component of that. So you’re going to be looking to new technologies, see how they can actually benefit you. And then this quote from I wouldn’t say my favorite coach and college football, but a coach in college football and Nick Saban, talking really about his mantra of trust the process, ensuring that the process that they have in place for their players, will lead to the outcomes that they want. So he says, Don’t think about winning the SEC championship, don’t think about the national championship. Think about what you need to do in this drill on this play in this moment. And that’s the process. Let’s think about that. What we can do today at the task at hand. That’s really you know, him saying that they’ve set up a process that will work. So you know, your team members, your teammates, all you have to do is think about your job at hand, and you’ll get the results that you want.
Why be Process Centric
So as I’ve kind of alluded to process, mining is really going to help you in a number of ways achieve a process centric view of your organization. First and foremost, it’s going to allow you to see kind of that, that digital factory for digital processes, of course, aren’t as easily observed as that physical assembly line. But process mining really helps you to see what’s occurring in your digital processes in your back office in your front office. From the end to end perspective, oftentimes we were only you know, privy to you know, how many invoices we paid out in a given month or what you know, how many sales orders we closed in a certain month, but we don’t necessarily see what led to those outputs. Process money is going to give you that because then you can see exactly what happened from start to end of your process.
I also like to think of process mining as kind of a game film for your entity and processes. It allows you to see you know what happened? What you did wrong, what you can maybe correct in the next in the next go. And it also really is a way to kind of create a playing field for your kind of friendly, friendly process competitions. Andy Grove, makes in his book high output management, he makes a comparison of, you know, great companies creating sort of a playing field for their team members to encourage kind of friendly competition and increase motivation. Process mining can really help with that, because you’ve got a real time Look at the playing field of your back office or front office process processes. And then it also really enables your managers to provide real time coaching to the team, which is sort of the other element of management that Andy Grove again, kind of sees as being the key to management is teaching and coaching your team as you go. And finally, you’re really going to allow your team to see how their actions affect their your global corporate objectives. So you can quickly see, you know, if I’m taking too long and a particular incident in ServiceNow, or I’ve not met my deadlines for paying invoices, you can right away see sort of the how that affects your overall objectives at a corporate level. And it also again, kind of gets back to your measurable impact across the three, three E’s of process improvement.
So that’s really how I see process planning, you know, beyond really just an automation tool, tool for finding automation opportunities can really be used as a management tool, and a corporate alignment tool to help you achieve your goals. So once you’ve implemented process mining, we often see, you know, really, companies enter into this cycle of kind of Process Excellence, which we call the Process Excellence flywheel. And this flywheel is something that kind of continues to build on itself as you as you progress through it. So first and foremost, after you’ve connected process mining to your process, you’re going to start to discover things about your process, you did not know we’re even there. So you’ll find process deviations, bottlenecks, rework, these are all opportunities really for process improvement. And so we can then take those and start to understand, you know, what are really the value opportunities that we would have by correcting those particular deviations or bottlenecks, we assess really then kind of the value of that and then prioritize what we’re going to get done to improve our process.
Second, or next, really enhance your process, whether that’s through coaching, your your team, standardizing your process, optimizing it in some way, or even automating it. And so that’s when you’re actually making changes to the process and implementing what you believe will impact your process. Next, you’re going to see, okay, what are the results of that enhancement? Did we improve the process like we had hoped to, and really going to be able to react then, and Rudy will talk about this today. But you’re there’s a whole new module within process planning that allows you to monitor those processes and then really react in in near real time to process deviations are things that you may need to, to act on quickly to get that process under control. So you should really start to monitor and react to your process via process lighting. And then finally, you’re going to go back and refine your process mining analysis, whether that’s adding more events to your event log. So you can get again, a wider view of your process, customizing metrics, or just overall kind of going deeper on your analysis of your process, which is going to in turn lead to more discoveries, which will then again, start the whole cycle over here. So that’s really how we see you know, using process mining in kind of a continuous way to really understand how your process is working and how you can continuously improve on it. So I want to do a quick demo here now of a ServiceNow instance, where we’re looking at sort of a tech support company using ServiceNow. So I’ll talk you know, I want to set the stage here in terms of what the businesses were, we’ll take the perspective of sort of a medium sized tech services company with multiple satellite locations for kind of customer support. And they’re basically looking at, we’re looking at our customer service and tech support processes, managed via ServiceNow. And really, our corporate objective, of course, is really to increase customer satisfaction and retention through our exceptional customer service. So we’ll take a take a look here, and we’ll take a look at this demo.
DEMO: Tech Support on ServiceNow
Tyler Safranek
All right, great. So here’s our process mining dashboard. And looking at our overall process here. First and foremost, the things that you’ll notice are we’ve got roughly 4000 cases, a case, in process mining represents sort of your, your process unit. And so that’s what’s flowing through the process. We’ve got 20,000 events, excuse me, let me back up. So our cases, here are the incidents that are in our ServiceNow instance. So we’re looking at a single month of May 2019, we’ve got 4000 incidents, we’ve got 20,000 events, and 18 different activities, 400 different ways that that those activities actually get done in 197 different users.
So I’m going to take the perspective here of really a manager of a one of these satellite locations, and I want to make sure that you know, my location is meeting, SLA is in terms of incident resolution, and then understand sort of what we can do to make that better. So I’m going to jump right away to my conformance tab, where I’ve actually set up tags to monitor SLA violations, as well as things that are maybe undesired sort of business process elements. So here, I can go and I’m going to change my pivot table to location, so I can see all the different locations that are being monitored. And then I can see my specific location, which would be location 93. Here, I can see I’m doing pretty well in compared to the rest of the other locations in terms of how many tags I’ve been I’ve been tagged with. So I want to drive my tag number down, that’s going to mean that I’ve got fewer SLA violations and fewer sort of inefficiencies in my process. And so I’m going to just pick on location 93 for right now. And then I’m going to break that out even further by my team. So I can see exactly how many tags are my team members triggering in terms of what what’s going on there and figure out maybe who needs some extra coaching or some extra help to really meet those SLA s. And so right now, I can quickly kind of see that my resolver number 15 is carrying a lot of the load in terms of tags. So let’s see exactly what tags he’s violating, looks like 15 violations in an SLA, and had to reassign the ticket multiple times, and seven, seven different instances. So that could mean a couple of things. Either the people in front of this resolver are not understanding that this guy is the guy to go to resolve these types of issues. Or there’s just there’s some sort of process problem here where we’re not getting the ticket to the right person right away.
So we can kind of break that down even further to understand maybe there’s a specific category where these violations are occurring, or there’s multiple reassignments. So I’m going to filter even further here on my multiple reassignment area to see looks like maybe category 53. So this is this is an opportunity now to look at it with the whole team to say, okay, category 53, whatever that may be some sort of specific tech issue, this category 53, should maybe come to resolver 15, right away rather than being passed around across your across your organization multiple times. And so that’s kind of just a quick view of maybe a coaching tip that you can give to your team to quickly kind of remedy these sorts of issues. And I think I’m going to call it there, because we’ve got plenty more to get through today. And I want to leave plenty of time for my co presenters here. So I’m going to hand it over now to Rudy, to talk a little bit more about the UiPath roadmap, and where we want to go with it. So really, I’ll run the slideshow for you if you want to just let me know when you want to switch slides.
The UiPath Process Mining Roadmap
Rudy Kuhn
So guys, before we jump into the roadmap, I would actually like to talk about you know about all hype automation platform. You know, the way we look at process mining for us, really process mining is not it’s not it’s nothing that should be isolated, but it’s really a very important part of our heartbeat hype automation platform. So before we go joined UiPath, in October of last year, there were only three boxes, you know, the blue one build, manage, and the red run box. So without any doubt, UiPath is the leading company when it comes to building, managing and running are attended and attended, or new test spots really great stuff. But as Tyler pointed out, you know, many customers are struggling with the question, okay, what kind of process Do we have, or how are the processes run, what should be automated in what order and what’s really going on. And that’s why UiPath acquired process goat along with another company. And right now, you know, we have more than just one product, we have actually four different products in the discovery box, we have automation of automation that basically is like a social network where everybody who is who is actively involved in the process or working in the company can propose automation ideas, we have test capture, test capture is like a like a little agent sitting on your desktop and you tell the agent, you know, to actively look over your shoulder, while you are executing a task or process it will record whatever you do, it will take screenshots. After you finish, it will give you a complete process definition document, which makes it much easier for the people to automate. And you can export, you know, this, this recording, like a macro already to our studio and, you know, compliment the bot with some AI or some document understanding and really get it up and running very fast. And then we have processed mining and task mining. And the difference between the two pro mining products really is that process mining is using data from the backend systems, like SAP or any other, you know, the big milestones. And tasks mining is rather looking into the front end of applications. Because if you think about systems like work law, you know, like for example, x or email or PDF.
Unfortunately, these applications are not creating any digital footprints we can really use for process mining. But if you think about the process, then these applications are also involved in the process. So that’s why we are you know, taking data from the backend. And from the front end, we are combining them. And I will show you know how this will look like in the future. So we better understand the entire process end to end, so we can better automate. But we don’t stop there. I mean, in your flywheel you talked about using process mining also for the measure point, because of course, after we have, after we have improved the process, and you know, let’s be straight, I mean, RPA is just another way of improving the process. You know, as a former consultants, I can think of many other ways our process can be improved. But RP is a great one. But of course, we want to understand the impact of RPA of RPA on the process. So this is why we use process mining in the measure part, as you can see at the end, in the blue box. And with our latest release, now we can even, you know, again, connect process mining with our action center, because we don’t want to stop with observing that something is going wrong, we really want to react on it. So for me, you know, the perfect combination of the tools really is processed binding for the diagnosis, and RPA as a treatment and only if diagnosis and treatment really comes together in the right way you can be successful in the long run in the long run. This is now really focusing on the positioning of process mining. So we start process mining for the understanding of the process of the real processes, because there’s a difference between where people will tell you in a workshop and what the process really is. I can tell you from experience, and everybody’s smiling, so you know what I’m talking about. And, you know, after we understand what, what, what’s really happening, we can automate, then we can monitor, and then we can react on it. And yeah, I don’t know if we have to tie but I know I can show you how this reaction actually looks like what’s how this really works. I think for that I need to share my screen.
Alright, so what you have, you have already seen our application, you know, it’s browser based, we run on premise we run or completely on, on the cloud, whatever you prefer. So this is a very simple invoice approval process. You know, the full complexities of the happy path actually looks like this very simple. Receive invoice, process invoice, fine check of invoice, then of course, we should approve, approve the invoice and then at the end, we pay the invoice. And if we add more activities, we can see what else is going on here. And if we add all edges, we will discover that we might have some compliance issues because some cases are bypassing the approval step.
But right now I’m looking at all data from October. And I have a bookmark here really to focus only in the last day of October. And on the last day of October we discover we can see We have one case, where the approval step, the mandatory approval step, for some reason was bypassed. And that’s not okay, you know that that shouldn’t happen. And that’s a difference on how we transform how we work with the data. So with our RPA, or with our process mining platform, we connect directly to the sources, we extract or we read the data and we perform this whole data transformation internally. And you are already talking about the conformance checking the tagging. So we can use for example, once we identify any type of issues with tagging, like, for example, approval is missing. Now we can trigger directly through our action center, we can notify the people or we can trigger some bots. And let me show you how this works. But I just realized for that, I need to log into my to my orchestrator, so we can see the outcome. So yeah, you know, this is live.
Okay, I am now in my action center. And yeah, that’s the that’s the action list the to do list. I really like because it’s empty. So nothing to do. That’s great. Now, I go back into my demo. And I you know, because typically, we transform the data every night or every hour, it depends on your use case. But you know, we don’t have the time to wait until tomorrow. So what I can do what I would do, I go to a specific page I set up, and we have this create queue Item button. So you know, normally nobody would push this button, this would happen automatically. Every time we transform the data, we update the data set. But for presentation purposes, I can trigger this now manually. And when I click on it, you know, the system now created this action item. And when I go back into my orchestrator, and I refresh my page, I will see that a few seconds ago, you know, when I push the button, this one item was created, and the title is invoice approval missing. Oh, that’s not good. So what can we do about it? You know, when I click on it, first, the system will show me some of the data, the master data from the case we just discovered, so we can see the case ID who is responsible for it. It’s only a small invoice and I know the supplier. So yeah, that basically that that’s fine. But still, before I approve it, I want to see the invoice.
So this is just a dummy, you know, just for presentation purposes, I can click on Open in web or this could be an invoice document management system. So I can see the invoice and say, Hey, okay, that’s fine, we can pay that one. So I go back, and I if I click on Open SAP, then nothing happens. Oh, yeah, right. Yeah, I’m not a manager, yes. So I need to assign this task to myself, so I can work on it. Now the buttons are activated, you know. And if I would click on Open in SAP, then basically SAP would come up, I would see this case, and I can do this approval manually. But again, I want to save this time, because that’s just boring to do that. So what I can do is, I can simply click on approve, and if I click on approve, I will trigger a sent a bot actually to act to, to approve this invoice for me in my name Michael with my credentials, I do that. And all I need to do is not to confirm that this action is completed. Now I do that, I come back to my list and it’s empty again. That’s what I like. Well, unfortunately, we don’t see that, you know, the update in our process money because this update will first we need to transform the data and run it again. But tomorrow, you know, we will see that probably after we paid the invoice. We approved the invoice afterwards. But still, the data is correct. Yeah, and that’s, you know, one way how we are integrating our process mining and functionality in the in the whole automation platform. Because we really believe you know that we need diagnosis, we need treatment. And we need to empower everybody who is actively involved in the process with a functionality like that. So if you are responsible for some cases, you simply deserve to know what’s going on. Because only if our employees are informed, you know, they can they will be motivated, they can have a positive impact on the process. So that’s why we make this technology available for everybody who is really actively working and involved in the process was just a quick presentation on how we integrate.
So the next slide will show you that, for example, we have these. We have these dashboards for example of SLA violations and if I see that we have 130 SLA violations. I mean, that’s great that we were able to reuse the numbers. But in future, we don’t want to see any SLA violations. So not too fast. So what we can use what we can do, you know, we can use the tagging and whatever functionality we have really to notify people in the Action Center to automatically trigger bots to take care of the problem. So this is how we close the loop, you know, from understanding, automating, monitoring, and down to reacting. But one thing we are working on for the future, and that’s the next slide. You know, we really want to combine process mining and task money, because only if you understand how processes are supported in the backend, and in the front end, you can really, you know, automate them.
We can see that on the left, we have the process and based on the dark, you know, shading of the color, you can see that process invoice, for example, in this in this case, when this process is our bottleneck. So we really need to speed it up. And before we can speed it up, we need to first need to understand how process invoice is really run. So you see on the on the on the right, please select a task on the left to visualize its execution. So Tyler, if you please click on process invoice, right, the system will show us actually, you know, process invoices, the big milestone the big task, and now we see the sequence of actions what is happening in the front end. So we are visualizing the back end process and the front end execution in one application. So we see that there are actually two different variants or two different ways how process invoice can be done. So that’s the first one you see the screenshots you see the applications, but there’s a second variant. So if you can please click on very number two, the system will show us that, you know, you know these are fewer cases, but the process is much more efficient. And if this is what I want this, if this is what I you know, the bot I would like to create what I can do, I can simply click on exporters to you it’s on the left on the on the right bottom on the right bottom, this little icon and the system don’t do it now. So it would it would actually export the skeleton for this bot directly to the studio. Again, I can start to build my bot, you know, by adding more functionality, maybe a little bit more Intelligence document understanding water. So this is on our roadmap, you know, we have process mining, it works, it works fine for a couple of years, we have test mining will be released soon as January wearability. But really, in future, we are looking into an integration of other tools. And this is just one of the items of the big items on our on our roadmap for better understanding and for better management and monitoring of processes. Awesome.
Brandi Corbello:
So I’m just going to quickly talk about Cushman Wakefield, hyper automation journey. And then I’m going turn it over to Salaman, who’s really been kind of the leader in the force, kind of behind our promises mining journey, and really executing that in our first use case with Tyler, and Rudy. So really, you know, like, if I rewind on, on where we started, our hyper automation program started last April, and really doing a proof of concept with our business. And they were excited, they approved that business case, and it was really RPA focused. So then kind of going down the year we, we started doing some education throughout the organization on what hyper automation was, and how it could help the business which continued to expand our business case, coming out of that we really needed to engage with a strategic partner. So that’s when we got engaged with Ashling. And standing up, you know, a partner to help us scale as well as think about, you know, how do we how do we scale you know, not new, but then also maintain what’s in production or what’s going to be going into production as we as we grow and velocity upticks. So that was at the back half of last year. And then really, you know, we had to define what a center of excellence looked like and what that meant. And we kind of closed out what is framework and governance look like? And that was really to support scale, and velocity as well as like, what are what does our infrastructure look like and how do we ensure that we can, we can support acceleration, so that was like back half of last year front end of this year. Now, once that was stood up, automation kind of started taking off for us? I think we keep actually busy, we keep ourselves busy. But then, you know, then the conversation has now turned into the second half of this year is how do we how do we broaden this, right? It’s not just about RPA, it’s about process. And, you know, we’ve got a few exciting things going on.
So in July, we did our first process mining pilot with a smaller process just to see like how it would work and like, what it would look like, and how we could really put a strategy together around that. And then right now Simon and Tyler have been working on putting a bigger process in our organization through kind of process mining, and Solomon is going to go into that a little bit, it’s actually a really exciting one for us in the commercial real estate space. And, and it’s, it’s really impactful, you know, across our organization, and even in our industry. So I think it’s a good use case. And it’s probably different, which is exciting, and has a lot of opportunity inside of it. You know, I think even part of this too, is we’ve started bringing in machine learning. So I think about all the different things we’ve brought into the program and how they all interlay together, and it’s been really exciting to watch, like the puzzle pieces kind of come together and watch, watch the program broaden as we really look at process mining, as a part of that, as well. So, so our program has quickly aligned to our organization’s strategic goals, which kind of foster exceptional client experience delivering operational excellence, focusing on people on culture, and then making sure we’re leveraging data and analytics to make more informed decisions. So all of these things that we’re talking about kind of feed nicely until all of the strategic objectives of Kushman. But our objectives were process mining, as we thought about the strategic lens and how we want to broaden it, you know, it’s an evidence based approach, as we as we look at the processes that support, you know, revenue in our business, it really, you know, focuses on operational excellence by you know, I think Rudy and Tyler both mentioned this, the business things, the process works one way, but you actually quickly learn, it probably could work 1000 different ways.
So how do you encourage operational excellence through like the use of the insights that you gained from this, and then really, you know, continuous improvement? I think the bigger thing for us and I’ve talked to Salaman about this all the time is, how do we monitor continuous improvement, you know, we’re going to see where we need change. And then you know, the biggest thing coming out of this is we’ll actually see that improvement was made, because we do have that evidence kind of behind it from the system. So I am going to turn it over to Salaman, who’s going to go through our use case, but high level just wanted the group to understand our journey, and you know, where process mining fits, and we’re really excited.
Salamn Khan
I’m a senior manager within the Cushman Wakefield, Intelligent Automation transformation team under Brandi Corbello. I’m going to talk a little bit about our use case that that we’ve partnered with UiPath and nationally to help deliver around deal management. And so from a background perspective, deal management, it really encapsulates the entire real estate deal transaction. You know, in the case of a lease, inclusive of the it’s all inclusive of really like the bid pipeline, phase deal execution all the way through to closing and reporting within a CRM system. So in our case, Salesforce, and so as we embarked on our review, in our assessment of deal manager, we found that it was a prime candidate for the process, right. So it had a very large user base user population in excess of 1000 ft. It was a across multiple service lines, roles and geographies. And it was fairly fragmented from that standpoint. And, you know, as you can see, as well, we had very high aggregate support costs, to help support the entirety of the lifecycle this process. And so, you know, during randomized discussions with appropriate business process, SMEs, we receive feedback, there’s a lot of different issues that plagued our process, right? They primarily centered around process delays, concurrent duplicative efforts in just limited insight and transparency into the entire process lifecycle, you know, if that wasn’t enough for our key issues, you know, the Salesforce system data also presented some challenges and that some noise existed in the Salesforce system data due to inconsistent deal entry. And what I mean by that is that folks are inputting deals in different stages during our pipeline process with really limited discipline. So, you know, this data quality issue coupled with very limited current state operational KPIs were hurdles that we were going to need to overcome. And so, you know, with all of this information, we knew that issues existed with this process, but we didn’t know exactly where to begin with it.
So a traditional approach would be to, you know, conduct, you know, extensive current state process discovery sessions across all of the different stakeholders, the different service lines, but when you start to weigh the benefit of that again, The option of using a tool like process mining, it becomes almost a no brainer, right. So using the process mining platform, we’re able to distill those objective process insights that Brandi and Tyler and Rudy have talked about in a very short period of time, using really only the tool, you know, a high level process understanding and connecting into our system, and in this case, using the Salesforce system logs. So you know, how did we begin, you can kind of take a look, and you can see we have a high level process flow above and is really on defining our scope, you know, you can see that we actually have nine different deal management stages within Salesforce. And as I alluded to earlier with the variation in how users interact with deal entry, we found that, you know, in order to mitigate this issue, let’s focus on a subset where we believe, based upon policy and business stakeholder feedback, that all deals are going in, or have to be in the system, at least in stage four. And so, you know, little did we know that this assumption was not entirely true, but we’ll get to more of that a little bit later. And so, you know, once we have identified that, that particular scope, we really wanted to, we really needed to work with multiple parties.
So Tyler helped us engage with, you know, our IT department, we were working in concert with UiPath, in really trying to understand from a very high level, the process, you know, fulfill all of those technical requirements and infrastructure requirements upfront, connect to Salesforce and extract the data and transform it. And when I say we did that, I really mean Tyler was doing it. So he’s done a fantastic job with it. But then really, once we’re able to extract that data, that fun, then the fun begins for our team where we can start going in and forming hypotheses and reviewing the output. But one really important point I want to, I want to emphasize here is that we were able to do all of this initial discovery without having to conduct those extensive and time consuming current state process workshops. So that was really invaluable and a significant benefit for us as a digital transformation team of being able to sell this to the business. You know, so once we had the data now what right, we start forming those hypotheses, and, you know, we’re taking a look at the different process variants, the different deviations from that happy path, we are, you know, identifying the associated parties and roles that are responsible for some of those variants responsible for some of those process delays or choke points that we see. And so it’s really challenging some of those initial assumptions that we had, you know, in reality, seeing throughput for a particular approval taking 10 days for a particular approval, versus what was expected was, you know, we’re expecting an outcome of less than a day for it. And you know, how I was talking a little bit about deal entry being required by stage four and progressing through? Well, you know, we discovered that that was false, right?
So we found a variant in the process where folks are actually going in and putting in stage six, which is, you know, which is a violation, one of our initial assumption, but more importantly, it’s a violation of one of our SLA s and one of our policies that we have internally. And this is something that we would not have been able to know without leveraging the process mining insights that we’re able to gather. So we equipped ourselves with these reality based, you know, metrics and process flows. And now we’re able to bring those insights to bear in conversations with the business process owners, which was incredibly powerful. And, you know, after you have those insights, what do you do with it? I think Rudy alluded to this earlier that, you know, you can gather all the data and all the analytics that you that you want on a process, but you need to be able to affect the able to impact it. And we in to do that we corroborated with the business process owners to identify what are those levers that we can push and pull? And, you know, what are the priorities that we have as a business? And so, you know, After determining what are, you know, what are those specific improvement opportunities? We, you know, we really, and this is to brandies point, we want to see what’s the performance lift or efficiency gain that we’re going to achieve from some of these lifts, right? You know, if we choose to automate components of that deal entry process, we want to see what impact it’s going to have on our data accuracy, right, we want to see that, you know, deals are being input and stages zero to three versus four to six. And so, you know, I think just kind of bring it all together, ultimately, in a matter of, you know, almost days, we were able to get these objective, comprehensive current state process insights, which, you know, traditionally, in traditional consulting, these are, these are things that take weeks and potentially even months given process complexity and the fragmented user base, you know, so, you know, with that, where are we going now?
We’re currently focused on building out a solution roadmap with some of these business stakeholders. You know, we’ve helped identify those recruitment opportunities and have it validated by the Salesforce data. You know, this isn’t, you know, we’re also focusing on you know, Brandi talked about alignment to enterprise objectives of continuous improvement and operational excellence and performance measurement is a big key to that, you know, we need to ensure that this is not a single point in time exercise, but an ongoing focus of understanding where we are today and where we’re going. Right. So, you know, what is the impact of our of our initiatives, right, be an automation via process improvement, bid training, implementation or policy adherence. We want to see how we’re doing against these measures. And so you know, we view process mining is really a platform that We’re looking to expand into other high profile high priority business cases across the enterprise. And, you know, it’s truly given us a window into our current state and really accelerated our process understanding without really being invasive into our standards, business operations. So this has been a very powerful tool that we have at our disposal now.
Panel Discussion
Marshall Seid
I’m going to go through some of the real time questions we’re getting streaming in right now. And then then we can go to some of the ones that were sent prior. So first question for the CW team. So Brandi and Solomon, it sounds like you were two years, roughly two years into your RPA journey? Do you have resident experts in your organization? Are you are you relying on a partner like ashling to do development scale? Like this is really about how long you’ve been on your journey? And then how are you building your internal competency as well?
Brandi Corbello
Yeah, I’ll take this one. So our journeys really only been a year long if you think about it. So we started last February, when I was more of an introduction and spending time, like just educating our executive team and getting buy in across the organization. You know, once we had that, and we had like, true business value that, you know, looked tangible in the business could you know, sign up for we, you know, we started, like, kind of building out the team now, you know, building out the team met, and actually was a part of this, you know, building our infrastructure to scale. So that was the first thing. The second thing, which a lot of people miss is what does your support model look like on day one, right? Like, people usually don’t start thinking about that until after year one. And we thought about that on day one. And we’re finding that we’re not having, like conflicting schedules, if you will, or like having to conflict priorities with our resources. Because of that. And I think that was a big deal for us. Our team internally is not developers, their team internally is actually processing analysts, and they’re tagged to business units. And it’s really helped us, you know, make sure that our business partners throughout the organization have somebody that knows their business from a process standpoint, but then can also interpret what the right solution is and be solution agnostic.
Marshall Sied
Maybe just to kind of add to your responses, because I’ve obviously got intimate experience in how Cushman has approached their Transformation Program. I think what you guys have done so effectively, is you’ve gotten the business constituents and stakeholders to kind of really lean in on things like process mining, right? And I think that’s, that’s powerful in its own right. But it’s allowed people to kind of feel like you’re sitting in their business, you know, their business, so you’re going to properly prescribe. And I really kept on using the term diagnosis, right. And the old saying goes is don’t prescribe until you’ve diagnosed or else that’s malpractice. And I think you guys have done a fantastic job. And that’s, that’s come through some of the trust that you guys have built through your business analyst team. So really appreciate that. The next question, I’m going to, I’m going to take one that was that was pre written, or preset. So this is also for Brandi and Solomon, and I think you guys saw but I think he did a pretty good job of puppet this one already. But how has this helped your automation program in the current state? And how do you view this bigger than just automation beyond automation and more transformational from a process mining capability perspective?
Salamn Khan
If I could draw from maybe one of those cliche consulting phrases that we used to throw around quite a bit, you know, we view process mining as the tip of the spear, right. So this is, you know, this is going to provide us with those deep process insights that we can take out to the business, right. So one of the challenges that you have, and I think brand has done a fantastic job of standing up this program, but it’s beating that inertia around change management, that you have to provide education and awareness on your different, you know, the different products within your technology stack, the value prop associated with it. For process mining, this is taking an existing process and beating that inertia that folks have with, you know, understanding what’s the value proposition, right. So, you know, you’re able to get those deep process insights, you’re able to really accelerate that speed to delivery and do this in a non invasive manner. I think that’s what makes it so exciting about the prospects that we have of deploying those additional use cases internally. I know that this is, you know, this is one where we’re making a foray into one of our specific broker market support Business, Business service functions. We’re looking To take this to other associated areas within the enterprise to but this provides, essentially that ammunition that you need in those conversations for, you know, for automation or for efficiency gains, and really lining back up to your enterprise objectives of operational excellence in ci.
Marshall Sied
So to comment on the sea level, just a quick response to one of the questions, yes, we will provide this, this video to everybody so that you can you can position this however you deem fit, however, it’s valuable. Another question for the CW team. So we got emptying the CW chamber here. Do you find yourself focusing more on larger end to end processes? Or you see, are you seeing this as more of a bottoms up approach for your organization? That got specific from a process mining perspective,
Brandi Corbello
I would say we’re looking at end to end and we’re focused on, you know, for us, like from a strategic lens processes that we spend a lot of resource time and money on, and are actually really meaningful to our bottom line, meaning they drive or support high margin for us. And what does that mean? Because for us, it’s, it’s about selling to the executive team, right, the executive team wants to see what are the things that are driving the most value as a process number one, and number two, how do we make them more efficient? Right, so have you gained more margin out of those processes? You know, I think bottoms up is an important part of a strategy, it’s not going to happen until I think we have our arms around the bigger processes that support our business.
Marshall Sied
Sure, well, it’s an outcome driven approach. I love it. A really I’ve got one for you here. So can you help the audience understand when you’re when you’re prescribing process mining over past mining? And I know that you kind of talked about the combination, but it seems like there there’s a lot of questions coming in about that, actually.
Rudy Kuhn
Yeah, you know, process mining really is the, let’s say, the more like the helicopter view. So you really look from high altitude, you really look at the entire process. Based on the big milestones you find in the systems and tasks, mining properties, more like the magnifying glass, if you really want no one to know that the biggest details are the process, the steps, the key actions that actually create an activity, you know, in the highest possible resolution and details. So, you know, for the big overview, process mining is great. So we’ve lost money, you can basically discover what should be automated. And when you take task binding or task capture, you look really close, sir, to the to the to the activities, then you can you can find out how to automate. So process mining, what to automate and test mining, how to automate and the combination really, you know, brings the biggest value.
Marshall Sied
So from a potential use case perspective, maybe leaning on your three n days, what are some other use cases that are more horizontal that a lot of other large enterprises can learn from?
Tyler Safranek
Yeah, of course, the big ones are ordered a cash purchase to pay accounts payable, accounts receivable. Some of the other ones, though, that I think are really powerful, and ones that we kind of use it through our production, execution and production planning, really helping to make sure that your planners are setting the right schedules for your production execution. If you’re manufacturers, of course, we also looked a bit at sort of deal management and point of sale, a lot of times we needed to collect point of sale data and make, you know, chargebacks, to our distributors. And that was a really powerful model that we use to really understand how that process was working on which kind of workload really, that whole team had to help kind of alleviate some of those processes. So I mean, supply chain is rife with opportunities for process mining. And I’ll say again, financial, financial, financial sectors to credit applications, banking, there’s all kinds of processes where we look that can really be mined and useful and process finding.
Rudy Kuhn
I would like to chime in here, you know, you can think about process in two different ways, core processes and support processes, and digitize processes, or fully digitized processes or physical processes with data attached to it. You know, if you think about a company like Porsche, if you heard about them, you know, here in Germany, they build these nice cars. So it’s, it’s a production process, they have to move around a lot of atoms to make money and but there’s a lot of data attached to it. But still, it’s a physical process. But if you think about, let’s say, a trading process, or know your customer process in a bank, it’s all data If it’s all data, there is a lot of you know, content and, and really, you know, food for us and food for personal and it’s a lot of data we can work with. So if you find core processes and these core processes are fully digitized, that’s the you know, that’s, that’s perfect, then we can deliver the highest value. And of course, the we need to have data without data, there’s Unfortunately, no process mining.
Marshall Sied
So just in general, what does the future hold for process mining?
Rudy Kuhn
Well, from our point of view, really a very, very seamless integration into our hype automation platform, I really believe that every process that is digitally supported in every system that is supporting processes in a digital way, one day will be will be analyzed with process mining, because it’s the only way really to get the transparency we need, the only way we can react on on any type of issues. And you know, the traditional way I was doing for many years talking to people, we will not replace it, but we will completely change the way we run this these workshops. You know, in the early days, I asked people, please tell me, how is the process run? Today, I walk back into the same workshop and I read know what the process is. And I asked the people, why is it like that, you know, from how to why. And it’s a, it’s a totally different, totally different story you have so what we see for the future, very seamless integration, we are moving into more, you know, taking the data, doing process simulation, maybe maybe even redefining the processes before we automate. So there will be more, more of that. And also, you know, stuff like predictive analytics, because if we know from the past, what can go wrong, then we can we can already predict the outcome of certain processes or cases before, you know, the sorry for my for my French before the shit hits the fan. So we can see, okay, this will never be finished in time. And we want to wait, we don’t want to wait until it’s too late. We can react, you know, in time and prevent the problem.
Marshall Sied
So Tyler, what is what is the future hold from your perspective?
Tyler Safranek
Yeah, no, I totally agree with a lot of what Rudy said, you know, it’s going to be tighter and tighter, sort of integrated with your process in terms of being able to react in real time. And I see, you know, really process mining, integrating even more with machine learning and making predictions. I think process mining exposes data sets in a way that are sort of unique, and that can really I don’t think they believe in leverage dead in machine learning, looking sort of process order and process timing, to make predictions and about what what will happen in an open case of your process. So I really see that being a huge opportunity. And a huge I think, direction for process waiting to go is really pairing again with that machine learning and artificial intelligence to make predictions about your process before you have an issue at the output.
Marshall Seid
So just to give Solomon and Brandi kind of opportunity to answer that question. I know you guys refer to that. And then then we’ll, we’ll call it a day. Brandi, What do you what do you think the future holds for the process mining capability, not just at Kushman, but I guess in the market?
Brandi Corbello
Yeah, I mean, I think I’ll add on to like, what Tyler was saying, you know, we’re starting to bring machine learning models, even into our process automation that will, you know, was traditionally rule based, right? So how do you make that, you know, more dynamic using machine learning and being just a little bit more predictable, or, or more dynamic and decision making? So, you know, to Tyler’s point, I do see machine learning, you know, kind of playing a role to like, give us predictability about our business. And, you know, you know, if this fails, right, like, what does that mean, not just to the process, but to the organization, right? What does that mean to the bottom line, if this fails, or if it, you know, up ticks and we can’t, we don’t have the capacity to like, maintain, right, what does that mean to our bottom line, I think is going to be really important as you look forward. You know, I think, I think a bigger thing, you know, for for services companies, you know, or the services industry. I think, product has always been really good at KPIs and measuring operational performance, but I’d say Services has always been lacking, right? Like, there’s hardly KPIs so how do you actually think about your true operations in the services industry and you know, given the current environment and kind of coming out of it, really starting to measure yourself on services and operations to provide those services I think will be a huge as well.
Salamn Khan
Yeah, certainly no, I, I’m kind of going to be parroting and echoing what Tyler Rudy and Brandi have covered off, right, but you know, more tightly coupled within our Intelligent Automation program and really look right at our core, we are focused on process improvement within the digital space. Right. This is I view this as a staple tool within our toolkit, right. This is a feeder into associated process improvement initiatives, but also within our automation pipeline, specifically, right. And the real eye opener for me is really been the speed to value that we’ve gotten through process funding. So I’m, you know, I’m looking I’m looking forward to like brand new setup, looking forward, ready to the additional releases and the enhancements that are coming. But no, I think this is this is certainly going to be a foundational component within digital transformation programs moving forward.
Marshall Seid
Thanks, Solomon. I appreciate it. And thanks for everybody for joining. Appreciate you holding on for a couple additional minutes. We will send out some information try to answer any questions we didn’t get to today. But yeah, stay in touch with us. We’ll continue to do more of these sessions as this fast paced, industry is quickly evolving. So, appreciate the panelists for joining appreciate all the attendees for attending.
