Podcast: Focus less on numbers, more on execution


Megan Wright
02/12/2018

When it comes to automation, the focus should be on how good execution is—not how many bots you have

rpa quality hero
Photo by Mike Kononov on Unsplash

“If it makes sense to have multiple robots, then you make multiple robots. If it doesn’t make sense, you just have one.” This has been Martin Ax’s key takeaway in implementing RPA and automation across his organization.

“I think people are too occupied with, ‘Oh, how many do you have?’ and it’s not really important. I think the important part is how good are you at developing them? How much benefit do they give? Is it part of the strategy?” he explained to AI Network Podcast host, Seth Adler.

It’s also about ensuring the quality of the automation is up to scratch, especially in an industry as highly regulated as banking. High quality automation should be a focus for all, Ax explained; “if not, you risk sabotaging yourself.”

“It would have taken two man years to update the process in question—it wasn’t a really good business case. But with robotics, we could do it in a week.”

Tune in as Ax and Adler discuss how automation can be a journey to the front of the customer experience, why no process is too small for those six sigma-minded folks, and what it means to automate in a safe and reliable way.

Listen now: 

[transcript]

Seth Adler: Martin Ax joins us. First, some supporters to thank, and thank you for listening. This episode is supported by the AIIA Network, the AI and Intelligent Automation Network is an online community focused on building the intelligent enterprise. Content covers a broad range of issues including digital disruption and transformation, task and robotic process automation, augmented intelligence, machine learning, and cognitive computing. Our goal is to help businesses apply these technologies, and built the intelligent enterprise of the future. Go to aiia.net to join.

This episode is also supported by RPA and AI Week 2017, the world decision makers and doers in process excellence and shared services meet in London this November to collaborate on the direction of task automation and augmented intelligence, share best practice, and discover strategies, tactics and initiatives, which industry leaders are already implementing for business success. 2017 is our second year of bringing this growing and exciting industry together. Go to rpaandaisummit.com for more.

Martin Ax joins us from the RPA and AI BFSI summit to urge his colleagues not to get too caught up in how many bots are working. He feels that the focus should be on how good you are developing automation, how much benefit you're receiving, and ensuring its part of an overall strategy. Automation could be a journey to the front of the customer experience once you're overtly effective. At automation development, no process should be too small for lean and six sigma-minded folks. Executing automation in a financial services environment forces you to be as precise as every one should be, based on the highly regulatory nature of the industry.

Welcome to the AI and Intelligent Automation Network on B2BIQ. I am your host, Seth Adler. Download episodes on aiia.net, or through our app on iTunes within the iTunes podcast app, in Google Play, or wherever you currently get your podcasts. Martin Ax.

Martin Ax: I'm very talkative.

Seth Adler: If that's fair, so am I.

Martin Ax: That's good.

Seth Adler: But, you backed your talkative up with it seemed like facts and figures, and maybe even a sense of what's going on.

Martin Ax: Oh, yeah, at least in my parts of the world.

Seth Adler: Right. When you say that you have knowledge in your part of the world, what do you mean?

Martin Ax: All my career, I've been in Denska. I've never been anywhere else, so I wouldn't assume to know what goes on in other places.

Seth Adler: I see.

Martin Ax: Having been on these workshops, these summits, I would say I have a good understanding on robotics and what makes sense, and what doesn't.

Seth Adler: Yeah, the number of bots that you have both is unimportant and also undisclosed.

Martin Ax: Yes.

Seth Adler: If, I remember correctly.

Martin Ax: Yeah, I think people are too occupied with, "Oh, how many do you have?" And it's not really important. I think the important part is how good are you at developing them? How much benefit do they give? Is it part of the strategy?

I think my bank has a very well documented strategy on where we want to go, and we are following that route. Development has been slow, because it's ... whenever you start trying to automate systems, you have to have your initial meeting. Sometimes, it's harder than to anticipate. But, we do have a lot of processes, around 100+ish automated, and-

Seth Adler: And you're saying process is automated.

Martin Ax: Yeah, process is, because one process can have multiple robots. It doesn't even matter. It matters how it's designed, and if it makes sense to have multiple robots, then you make multiple robots. If it doesn't make sense, you just have one.

Seth Adler: Yeah. I want to unpack a couple of things that you said. You have a well-defined strategy.

Martin Ax: Yes.

Seth Adler: Without giving away proprietary information, what could you share with your colleagues that are in the same boat as you, and need to be doing this, as far as the strategy…

Martin Ax: The bank is on a journey to focus on our customers. A part of that is, of course, to make us as reliable choice as possible. That means looking at all processes we have that is very heavy on manual labor. I think one of the strategies to go through them all, even the small ones, which then requires you to be very effective because any kind of IT development is costly.

So, choosing to do something you do once a year for an hour or two, you really have to be very effective of development before that makes sense. That's the strategy to go through it all, and to automate it, if possible, if it makes sense.

Seth Adler: I wouldn't imagine that would have been the first or second project, though, right?

Martin Ax: No.

Seth Adler: Understanding that you have 100 processes that have been automated, let's maybe go back to the beginning. You said it was slow-going to begin with, which makes sense. Number one, you've got to convince everyone. Number two, you've got to potentially clean data at that point, right?

Martin Ax: Yeah, and also again, it's a new technology, and getting resources trained. We trained everyone ourselves. Of course, with the standup material the vendor had, but still, we had to educate our own. We didn't get anything from outside, so it took time to do this. Also, setting up the environment within a banking industry is ... everything has to be correct. Everything has to be as it should be.

Seth Adler: There are higher powers-

Martin Ax: Oh, yes.

Seth Adler: That need to be pleased, right?

Martin Ax: Yeah, and still I think everyone should have a view on high quality on automations, because if you don't, then you can risk sabotaging yourself and also there was a lot of critics in the beginning on, is it safe? Is it risky? If you use it with care and consideration, and follow your internal procedures, then it is safe and reliable.

Seth Adler: These are the table stakes, what I was saying, it was in a highly regulated industry, there's even more.

Martin Ax: Yes.

Seth Adler: You said that you had a vendor. You also said that we didn't get anything from outside, so I wonder ... first, do you mind sharing who the vendor was, or is-

Martin Ax: We're using Blue Prism at the moment.

Seth Adler: What did you not get from outside, if we understand that you are using Blue Prism?

Martin Ax: From my point of view, going to an academy, or a school of some kind, and getting your staff trained, or even hiring in people who have this knowledge, was merely an option. We made a deliberate choice also. We chose five current developers we had, offered them to get this training, and then we trained them using the old piece of paper with things, and then we went on from there.

Seth Adler: When you say you identified five developers, how did you identify the talent? What did you ask them to prove, or what did they need to have on their tool belts?

Martin Ax: We already knew that we wanted to get started.

Seth Adler: Of course.

Martin Ax: So, it was about finding five people who were motivated, and how I could see fulfilling this role, because we use developers to do this robotics, even though the vendors, all of them, say that our product doesn't require fully specialized software developers to do this. I definitely saw a meaning of having them, because a real software developer is optic-oriented, and the reuse of these things are, I think, key for long-term payback.

 I found the five people I thought would be most open for this, and they were. It was an offer, and they accepted and wanted it. It was something new, it was something exciting, so win-win for everyone.

Seth Adler: Yeah. When you say that folks will tell you that you don't need in-house talent, it sounds like you're not necessarily talking about the tool belt. You're talking about the mindset of the internal program.

Martin Ax: Yeah, I think ... I always am a strong believer of having in-house resources, because it's an investment, and they are less likely to leave if they are consultants. It makes a lot of sense to build out in-house capabilities. I think most banks and insurance companies want to keep their talent within, so it makes a lot of sense.

 I would have liked to maybe have had maybe one or two guys from outside who had maybe two or three years of experience, that could get us up, because we did pay for all the mistakes, and we learned a lot. It was a very high learning curve in the beginning. I think a nice combination of hiring one or two, just to get you started, not to make all the stupid mistakes you can avoid.

Seth Adler: Right, of course. The beginning, we've been referencing it, when was it?

Martin Ax: Around November 16 ... actually, no, 15. Sorry.

Seth Adler: Yeah.

Martin Ax: So, around November. But, the bank has already ... we had different kinds of software even before it was called robotics. We had something called QTP from HP, which was developed around testing. We just used it for our production, instead.

Seth Adler: So, you cheated is what you're saying.

Martin Ax: Yeah, we re-purposed this software. Essentially, it's the same ... at this seminar, we talked a lot about also can you use robotics for testing. Of course, you can. It's to do repetitious work, and testing is just that.

Seth Adler: What was the proof of concept that you chose, or how did you arrive upon the proof of concept? I guess, both questions.

Martin Ax: I think we just chose a process back then that made sense, that we wanted to try. Something not too complicated, and yet had some benefits that would persuade ... even then, it was decided in all, we weren't doubting that this was a good idea, we just wanted to know was it feasible? Was it something our systems could handle? We just started.

Seth Adler: Something big enough, but not too risky, is that what I'm hearing?

Martin Ax: Yeah, exactly. I think one of the first processes we did was updating customer information with an email. Something where we had a large volume ... I think we had 80,000 customers.

Seth Adler: This was a customer-facing thing.

Martin Ax: Yeah, you can say it was a ... we had a third party company who had contacted some of our customers, and asked about are we allowed to give you market information. I replied yes or no, and then we would update the information about this.

Seth Adler: Mm-hmm (affirmative).

Martin Ax: I think with the volume of customers, it would have taken, I think, two man years to update this.

Seth Adler: Sure.

Martin Ax: It wasn't really a good business case, but with robotics, we could do it within a week, and then it didn't really matter. It was a good business case, but it was a hard task, so it made sense.

Seth Adler: Two years FTE done in two weeks.

Martin Ax: Yeah, but this, again, with all RPE, you have to think about the customer impact, what do you need? This was so simple. Just two field, yes or no for are we allowed to market, and the actual email address.

Seth Adler: Also, a task-

Martin Ax: Oh, yeah.

Seth Adler: That we simply ... no one wants to do it, and to have our people do it is almost frustrating that we have to even ask them to do it.

Martin Ax: Asking highly qualified bankers to sit and update these things doesn't make a lot of sense. But, then sometimes we have to have these things. It was a very good candidate for doing this.

Seth Adler: Following this proof of concept, how quickly did you ramp up? What were the next processes that you-

Martin Ax: To be honest, I don't really remember, because we started up heavy.

Seth Adler: Fair enough.

Martin Ax: We started with five people, and actually 10 analysts who worked ... I think it took some time before we actually started going beyond that. We invested, I would think, heavily in the beginning.

Seth Adler: How long was it until that team could hand however many more processes? Again, the number doesn't necessarily-

Martin Ax: I think it took us at least three or four months. It took us one month to learn the software, and then we started up. I think it took, I don't know, three or months. To be honest, things have gone very fast ever since-

Seth Adler: After that.

Martin Ax: Oh yeah. Yeah.

Seth Adler: At the beginning of the year, how many processes were automated, if six months later were at 100?

Martin Ax: I really don't remember.

Seth Adler: Ish.

Martin Ax: Also, just a thing I didn't mention-

Seth Adler: Yeah.

Martin Ax: Just after a few of these have been produced, we decided to launch a Center of Excellence. That also took some time and some energy from us, because when we started initially, we were the only one within Denska who were doing this. Of course, with the launch of the Center of Excellence, which was in our process department, yeah.

Seth Adler: That helped-

Martin Ax: That helped to get other factory teams up and running, and helped us improve our standouts in all of these things that should be in place.

Seth Adler: So, here I am, again, talking about tactics and specific numbers of things. You keep on telling me, "Well, it's really not about that," which I try to know that to begin with, but it's very easy to get caught up in numbers. Go ahead.

Martin Ax: I think, like everything else, make some decisions on where you want to be. It's just like normal development. Start with your requirements. Where is it that you want to have this robotics take you? From that, make some planning, hiring the people you need, and make sure you can deliver. I think that that's it.

 The amount of ... some robots will take you years to build, because they are so comprehensive, and so heavy on, you can say, regulatory requirements, so it will take long. Some robots can take as little as a day, depending on what you need and want. These numbers, I could say 50, and you would be happy because you have a number-

Seth Adler: Now I have a number, exactly.

Martin Ax: Yeah, but you can't use it for anything. There's no reference.

Seth Adler: I think for folks listening, the idea is benchmarking, and your point continues to be that that is not the way to benchmark.

Martin Ax: No, it's too simple. It doesn't tell you anything. I would say there's a linear movement for us, and we are improving quite considerably. The thing I am looking at is the average development time for each of these robots. Are we improving? You can say, on an average thing, if you average everything out, you should be able to see a positive move and definitely we see a very positive move from when we started to where we are now.

Seth Adler: What kind of metrics can you give us there, without, again, giving any proprietary information?

Martin Ax: I don't know if it's proprietary. I would just say we definitely more than halved our development time.

Seth Adler: Excellent.

Martin Ax: That's something you should see. Again, if you do this the right way, and you start doing this as objects, and you really treat it as IT, I think I'm one of the few people who really think this is IT.

Seth Adler: As opposed to business?

Martin Ax: As opposed to business. I am part of business, but we are IT as well. This is an IT tool. It's a lever that you can use, and we use IT developers for doing this. When you start using objects, and you can reuse, and reuse again, developing time on the entire processes go down. That's why you should see a decrease in development time, and reuse should go up.

Seth Adler: I have spoken with, now, more than a few folks who are engaged in automation. It does break down. It's IT. It's business. It's a combination of IT and business.

Martin Ax: Yeah.

Seth Adler: Why do you think you're one of the few people that wholly puts it in IT, which is a win for you?

Martin Ax: Again, as this is IT, and I am part of the business, it doesn't really matter. I think if you try to exclude IT, you will fail. If you try to exclude business, you will fail. If you take it as a whole, then you will succeed.

 If everyone is on board, then it's a lot easier, unless you are ... because we don't have to fight the battles along the way, because we are aligned that we want to do this. The question purely becomes, you know, do we have the processes for doing this? Do we have the right method? There, we have the Center of Excellence to guide us there.

Seth Adler: Absolutely. I want to get to that, but the last question on the first part is that you kept on saying "we were doing this", meaning my assumption is that you had sea level buy in from the jump.

Martin Ax: Yeah.

Seth Adler: How did you achieve that? Were you just a lucky guy?

Martin Ax: Well, we saw the software, and I went back to my director and I said, " This is really promising," and he trusted me and said, "If it is what you say it is, we can use internal resources. Let's test it out." When that became apparent that it was what it was, then you can say, "Senior management took a decision. Let's do this the right way, the Center of Excellence, or the best practices," and that made a lot of sense. I think we were very brave just to test it out and see was it worth all the hype? And, it was.

Seth Adler: And it was, which brings us to the Center of Excellence.

Martin Ax: Yeah.

Seth Adler: How did you go about creating that? When did you know that it was time?

Martin Ax: Well, this is the part I wasn't a part of.

Seth Adler: A-ha.

Martin Ax: I transferred knowledge into it, and I was, you can say, consulted during the build up, but it's not my department. It's a separate unit located in our process department. So, it was decided that this was the time, because the proof of concepts were successful, and senior management believed in it, and then, "Let's do it the right way."

 Playtime is over, now let's get serious. Let's scale up. So, that decision was made. We, of course, reported it, and it took, I think, two or three months to transfer knowledge from us into them, and then we became a factory team like everyone else.

Seth Adler: That's at two or three months to get the Center of Excellence going, or just your part of it?

Martin Ax: I think it took two or three months just to get them up and running. We already had the basic building blocks, the service set up, and the separations duties were already there. They took over, and started rolling out, and started improving the processes over there. Again, when you start something up, you can of course assume some things. When it becomes a reality, then there's room for continuous improvement just like everything else.

Seth Adler: Just like everything else. On our way to process excellence.

Martin Ax: Oh, yes.

Seth Adler: As it were. But, how do we have you, Martin, in this role to begin with? Let's go all the way back. Where are you from?

Martin Ax: I'm originally educated as a normal banker. I have a normal banking degree.

Seth Adler: A normal banker.

Martin Ax: Yeah.

Seth Adler: Your ordinary, average banker.

Martin Ax: Yeah, I started as a student and got my education through the bank, and then my career went from there to advisor to ... at some point, I actually went into Group ITS, a business developer, and then I went back to operations, and I was given the chance to take over the department of the IT people, and then I did. You can say the rest is history.

Seth Adler: Indeed. Has this been your only job out of college?

Martin Ax: Yeah, in Denmark, yeah. This has been the only thing I'm doing. I've been there for 19 years-

Seth Adler: Before that, were where you from?

Martin Ax: I was just a student. I went directly into Denska, and I stayed there. That's the, I would say, one of the big advantages of a large company like Denska. You can go multiple different directions. For me, I was just lucky that I proved myself during these different roles, and I was given opportunities along the way. I went into group IT, I've been in operations, I've been in the front, and I guess that's also one of the things that makes me a strong candidate for the automations, because I've seen it all.

 It's funny, I've also visited other places, other banks, other ... it's the same. For me, I always bog this down to there's no magic for robotics. It's all about data. You move data from one place to another, using different tools and, I think that what makes it simple for me is its just data.

Seth Adler: Right, it is just data.

Martin Ax: It is. There's no magic here. Many vendors will beg you to believe that it's magical, their product, but it isn't. It just copies data from one place to another in a highly sophisticated way.

Seth Adler: When looking for a vendor-

Martin Ax: Yeah.

Seth Adler: What advice do you have for your colleagues?

Martin Ax: I don't really know, because it's a thing. We chose Blue Prism to start with, a very strong vendor, and they still are. I guess like the rest of all the banks and insurance companies we're looking for the next thing. We want to use AI. We want to use cognitive. We want to use machine learning. Which product will that be? I think we want to make sure we're not completely dependent on any given vendor, but my advice would be just do some PUCs with multiple companies.

 We had some companies who failed completely, couldn't actually get their software to work, and Blue Prism could, and they were the most competitive in matching for what we wanted, so we used them.

Seth Adler: Keep your data.

Martin Ax: Oh, yeah. We have nothing outside the bank. Yeah. Keep looking, I would say. We are looking at multiple vendors, still, just to see and make sure we still have the thing that suits us the best. We are moving into another vendor, I can't say who, but we don't want to be dependent on anyone. We want to keep our strength, which is agility. This is also something RPA supports.

Seth Adler: Right. I have three final questions for you. I'll tell you what they are, and then I'll ask you them in order.

Martin Ax: That's good, because you have to repeat them, or I'll forget.

Seth Adler: Yeah, exactly. What has most surprised you at work, along the way? What has most surprised you in life? Then, on the soundtrack of your life, one track, one song, that's got to be on there. First things first, you said you've seen it all, right?

Martin Ax: Yeah.

Seth Adler: What has most surprised you at work?

Martin Ax: I don't really know what surprised me. Even in my young age, I think I consider myself a veteran, so nothing really surprises me, and I don't really remember anything surprising me. I guess what surprises me the most is how similar these problems are.

Seth Adler: To everything else.

Martin Ax: Yeah, and also you're going to say that different departments, different business areas: personal banking, business banking, C&I, and to be honest, it's the same problems. It's exactly the same. They think they're not the same. They think they're very unique to each of them, but they're not. It's all data.

Seth Adler: Were your folks kind of even-keeled, well-reasoned folks? You say you're not surprised,  is it because they prepared you well for life?

Martin Ax: I think the bank prepared me well for my work life.

Seth Adler: Fair enough.

Martin Ax: I think my father was a shop keeper, and it was his own shop. So, I learned to have respect for my own business, which Denska is, so yeah.

Seth Adler: Being the son of ... I've spoken to a number of folks, executives at large institutions whose parents were entrepreneurs.

Martin Ax: Yeah.

Seth Adler: You look at your desk as he looks at his shop, which is, "I'm going to be here in the morning. I'm going to turn off the lights at light, and no stone will remain unturned." What did he instill in you? When did you realize that there were lessons being instilled in you, I guess, in watching him to go work every single day?

Martin Ax: Well, you realize it when you start experiencing your work life afterwards, like real work life, that these are very sound principles, and they will take you far. I've seen many people not making a career, and I succeed. I guess, for me, a lot of these principles are very common sense. Treat your job with respect, and treat it as it is yours, and be happy. That's the thing, if you're motivated, and you love your work, you will go far. If you hate it, and you are just looking at the clock, you won't go far.

Seth Adler: That's no way to run a day, let alone a life, right?

Martin Ax: Yes.

Seth Adler: Which brings us, by the way, we spoke of your father. He was an Ax. You're an Ax, right? His father was an Ax.

Martin Ax: Actually, my mother ... it's my mother's family name.

Seth Adler: Oh, it's your mother's family name.

Martin Ax: Oh yeah. He was lucky enough to be granted permission to use it.

Seth Adler: Interesting. I see. Then, what is the story behind this fascinating last name?

Martin Ax: Like I told you, it was my great, great, great, great, great, great-grandfather, who made shoes, and was appointed to be the city area shoemaker. He wanted a unique last name, so he chose this name. We're actually not sure what last name he had before this. It's too far gone in time, but yeah. He chose a unique name, and it's followed us ever since.

Seth Adler: As far as those that argue for a work/life balance, take that. This guy changed his last name-

Martin Ax: Oh, yeah.

Seth Adler: So much that we don't even know what the other last name was, so that he could serve the community through his business.

Martin Ax: Yes.

Seth Adler: What's most surprised you in life?

Martin Ax: I guess being at the very tender age of 40, that don't take anything for granted, and take care of yourself, and live life to the fullest, because I've seen ... I've started to see some people I grew up with already not being here anymore, and yeah, don't take anything for granted.

Seth Adler: We're about the same age. It's nuts, right? I mean-

Martin Ax: Yeah, I don't meet anyone anymore who hasn't had someone close by, having cancer, or something else. People sometimes think things are fair. Life is never fair.

Seth Adler: Agreed.

Martin Ax: Don't assume that in 10 years from now everything will be better. Live life now, because ... that's my surprise. I thought I would live forever. Most likely, I will not.

Seth Adler: That's pretty much ... I think, I haven't done the math on it, but I'm pretty sure it's all of us that don't live forever, every single one.

Martin Ax: I think Walt Disney might live forever.

Seth Adler: We'll see, right?

Seth Adler: Let's give him time.

Martin Ax: Yeah.

Seth Adler: On the soundtrack of your life, one track, one song, that's got to be on there.

Martin Ax: I'm Walking on Sunshine.

Seth Adler: Look at that. That's from the 80's, you know?

Martin Ax: Oh, yeah.

Seth Adler: This is a very positive song.

Martin Ax: Oh, yeah.

Seth Adler: This follow suits with your philosophy.

Martin Ax: How can you not be happy when you hear that song?

Seth Adler: It's impossible.

Martin Ax: Yeah.

Seth Adler: Martin Ax, I appreciate it.

Martin Ax: You're welcome.

Seth Adler: We'll see you down the line.

Martin Ax: Hopefully so.

Seth Adler: And, there you have Martin Ax. "Like everything else, make some decisions on where you want to be. It's just like normal development. Start with your requirements. Where is it that you want robotics to take you? It's just data moving from one place to another in a highly sophisticated way." I very much appreciate Martin and his time. I very much appreciate you, and yours. Stay tuned.

OUR BENEFACTORS