RPA90X with Matt Gallo-Webcast Q&A Transcript

Edward Brooks: Hi Matt, first of all thank you very much for doing this. I’m going to break one of the rules I promised myself and do this from an airport, which is completely wrong!

Edward Brooks: One of the many reasons for doing these webcasts was to educate the marketplace and show people the interesting things that are happening.

Inevitably, there’s going to be a bit of a sales pitch involved in these things, but what I was fascinated with was what you were doing over the last couple of months… just this whole concept of boxing up RPA.

Everyone is keen to get started, everyone wants to move fast but it is how you avoid disaster, how you keep things moving, living up to the promise. That’s why I was fascinated to talk to you about it more.

Do you want to give a bit more information, just your background and what you did in the last couple of years?

Matt Gallo: Yeah, so Accelirate is actually the largest niche provider of implementing RPA in the country. This is really our only focus. A couple of years ago (as an example) I probably only had two employees in the US at the time.

I was group president in automation where we had a bigger footprint, but RPA in general was just really starting its wave. We looked at the market and said ‘hey there’s going to be a significant need for implementation partners that have a focus on getting these programs up and running, that can do it on-shore’, because a lot of the back office is here and without the heavy costs of the large management consulting companies, it’s all about the scaling and ROI.

You’ve got to have a balance between all of those different components to have a successful program. We started a couple of years ago, we are now over 80 employees and growing. By the beginning of the new year we already have things in place that are going to put us over 100, so it’s been very well received.

Edward Brooks: I don’t want to get bogged down in the offshore/onshore debate but the reality is most clients especially in the early stages want onshore because they can double your fee as well.

Matt Gallo: Yeah, so we talk to our clients about a progression, we always tell everyone that the first 90 days, especially usually at least 180 days, need to be onshore and often on-site where you’re right next to team members working together, helping to get ahead of any potential pitfalls before they occur, because executives are willing to. There’s tolerance for setting these programs up but they only last so long.

You can only keep going back to that well for so long, though we always say “Hey you know the first 90 days, six months, get it onshore, get it onsite, utilize a partner, get your program off the ground and then see where you want to go from there.”

Then look at other ways, but don’t start there winging it when you haven’t even dipped your toe in it.

Edward Brooks: Yeah, I think that’s the same thing everyone is asking for as well. The theory is there, there’s definitely progression but short term, sit side by side, stay in the same room.

Matt Gallo: The other interesting thing that we’ve seen, it seems counterproductive to what an implementer would want. We spend a lot of time focusing on building up our clients’ COE capabilities in-house for them because what we found was if they don’t have anyone, there’s no enthusiasm, there’s no interest, there’s no people, there’s no advocates, there’s nobody driving the program and it ends up just stalling because nobody is driving it on their side.

Whereas if we get them to build up their COE capabilities, there’s a lot more excitement and scale ends up growing exponentially.

Edward Brooks: I guess those were the kind of challenges that led to the program.

Matt Gallo: Yeah, absolutely, RPA90X in general was… the first year you’re obviously going to learn from your successes, you’re going to learn from your challenges. One of the things we saw was that there was really no failed program; they were just stalled programs, companies that were buying bots and then a year later still had two bots in production.

Or you know they’d have 20 or 30 bots in production but all of a sudden they didn’t have the right COE in place and the right governance structure across their different departments and all of a sudden code was different everywhere, maintenance programs were different, who was controlling what, nobody knew.

So a lot of times we’ve worked with a few enterprise clients that after six months to a year of getting these things in production had to hit the reset button and go all the way back.

And then lastly we also saw a lot of companies failing to set up proper budgets. RPA is being sold fast and sold hard as low cost, cheap, quick to implement and all those things are true.

Edward Brooks: You make it sound easy.

Matt Gallo: Yeah, exactly easy, anyone can do it, you can and while that resonates with a lot of people because it’s the magic bullet everyone’s trying to get, it doesn’t end up being real. Real clients need proper budget, get the right people to buy in, set the correct expectations, and make you really do need to have IT involved.

This is a business driven program, but without IT being involved, you’re going to have a lot of shortcomings on the implementation and again it’s not failing, it just leads to a stalled program.

Edward Brooks: Oh yeah, I mean politically it’s just suicide but we were just talking with one of our prospects in Asia, they are one of the large systems integrators and purposely did not involve IT and we are going ‘this is just madness’.

Matt Gallo: Yeah, it really is, one of the things we’re seeing especially when a program is new, people aren’t setting up a head of COE to begin. The business isn’t saying “Hey you’re in charge of this” and the other reason that people are reluctant to take that title is they’re afraid to leave their day job for this job thinking that the business hasn’t fully bought into because six months down the line if it fails, then what are they going to go back to?

So there’s a lot of people that are holding on to their old tasks because the business isn’t fully bought in and that’s why one of things we say is “buy-in” and that comes from a lot of different levels. That comes from upper management, the executive team, from IT, making sure that everyone’s bought into moving forward with this.

Now that doesn’t mean it’s a blank check and that’s why a strong but quick implementation strategy is incredibly important. Take 90 days, get it done right, get it in production, see the results, focus on things that are going to give you the right result.

Edward Brooks: So why 90 days? Is that because it’s kind of feasible, you deliver in that period because politically it is a size you can bite off quite easily or organizations can plan far ahead?

Matt Gallo: Yeah, there’s a statement that you can’t have a baby in one month with nine pregnant women and the reason for that is things just take time. Trying to force something in six weeks is just not realistic.

Everything takes time, everything takes planning and also in 90 days, the way we approach it, is to have three different resources doing three different things at the same time. One person is planning infrastructure, setting up a dev, setting up a production environment, another person is worrying about education and their valuation and doing process discovery and analysis.

The other one is really getting into the coding and then someone’s doing the testing and people’s schedules are always changing and you’re not going to have everyone available. So really we say it takes 90 days to get a program off the ground and into production.

Edward Brooks: I couldn’t agree more. I was trying to work out the 90 days, 4% is just technology what percentage is not technology?

Matt Gallo: Yeah, I don’t have that exact metric tracked, but it’s probably a healthy balance of them all to be honest. Each person in those first 90 days plays a pivotal role because you’re constantly trying to help the client educate and evaluate.

Because we don’t want to spend six months telling them everything they should do. That’s what a large management consulting company is going to do. Six months they’ll get a million dollar piece of paper that tells them all the wonderful things they should do and nothing is going to be in production and then the funding of itself.

But we feel that executives are really willing to spend the money on a quality 90 day program that’s going to give them a lot of value and results that they can attain and it also allows them to set a proper budget for the following six months.

Edward Brooks: And just in terms of the expectations over that kind of three month period, first of all, how do you manage the expectations based on what people have been told, in terms of RPA?

Matt Gallo: Yeah, that’s part of the reason we do this on-shore and on-site for these three months because you’re managing a lot of different people’s expectations and ultimately you really want to understand what their goal is.

Some people you can tell right away that they are off to the races, everyone’s bought in, they have 20 well-designed process design documents, they’ve already had lots of consulting and you just want to get heavy on the development and implementation.

So everyone’s a little bit different, typically what we see though is it depends on how well it’s bought in, how much IT support you really have. If IT is fully involved, the process design documents are well vetted by some- — and I say well vetted by somebody that understands RPA not just “Here is 20 processes we all hate and would love to get done, we never have to do the rest of my life”.

We look at them all and we are like “No, no, no, no”. And then they’re back to the drawing board. So a lot of times it is really resetting expectations and coming and really tell them “Hey you know maybe we can get one or two of these included as a part of the three to five processes that are in production but you really should just let us help you decide on which ones are going to bring the best ROI”

Edward Brooks: By the end of 90 days you reckon three to five- — this is 90 calendar days, this is not 90 business days. You can actually decide-

Matt Gallo: Yes, three to five processes in production. We work with the client obviously, so it’s going to depend a little bit on “Hey we would much rather get this heavy, very complex process into production than a few mediums or five simple ones”.

We try to do a balance of the few, so they’re seeing it from a variety of angles, that’s why we say three to five. We usually shoot for kind of like a high, medium, low as you will.

Edward Brooks: And then what? I’m going to guess after that 90 days what happens next? I mean are you expecting to continue the program for them or you are looking to make them self-sufficient or a bit both?

Matt Gallo: So what happens after RPA90X, a big portion of those 90 days is really helping the client understand their program with a 360 degree view. What we see a lot of times is, yes, now you’re set up, you have a decent governance structure in place, you have the right infrastructure, the right processes identified, team members are educated but what’s next?

And what we often see a lot of times is, if you’re not looking at this from a whole 360, you end up with one pitfall. We have a progression for our clients and we tell them “Hey some of this can- — if you want to move to lower cost off-shore” like I mentioned is something we help build out for our clients so that they can get the benefit of the lower cost but the same benefit of the good ROI from their program.

So it’s a progression, we believe that after setting up your environment, you need to actually build out a robust COE and then actually, there’s a lot of things when you’re looking at the cost of change, the cost of support, different predictive analytics monitoring.

At the end of the day all of these things were tasks that were monitored. You probably had dashboards; reports were generated by people performing them. Now also, all those things have gone away.

So you want to rebuild your capability around managemental workforce, it’s digital workforce, but it’s a workforce and it needs to be managed properly and you need to keep your KPIs in front of you, including; “How is this your infrastructure performing? How are the bots performing, what is your utilization?”

A lot of people forget about the utilization of their license. Actually without a proper COE, the utilization sits in the 50 to 60% range whereas with a proper COE, that’s sitting in the 85 to 90 range.

And for those that don’t know bot utilization is- — one bot can perform one process at a time, it’s not concurrently. So you get 24 hours of utilization, so if you are only using it for 12 hours out of the 24 then you’re at 50% utilization and if you then get another license your cost goes up from a software perspective and your infrastructure will end up going up as well.

So that’s why setting up a good COE- — then we go to what we call RPA 360 and that’s really a program around how much automation you really want to get done in the next year and then ultimately a managed robot operation center, ROC, where the maintenance occurs because the last thing we help our clients understand is, if you’re not careful ,you’re going to replace a $50,000 data entry person with a $120,000 RPA developer and all of a sudden in a year later, you’re going to go “Why haven’t my cost gone down?”

Well you replaced this with that and you still have your license, and of course you still have your infrastructure now you’ve just moved it all, but you haven’t actually improved it.

Edward Brooks: That’s the conversation that always gets lost, of replacing your low cost job with a high cost job, add up all the cost, the infrastructure cost and the consultants on board, people seem to forget about that and rush to do stuff.

Matt Gallo: Absolutely, actually it’s funny because I think that’s the part of the next slide is-

Edward Brooks: Let me just ask you something, I was surprised that without this it was even as high as 50% utilization, for a bot I would have guessed it would be in the 20s.

Matt Gallo: At 50% that’s an average, obviously we do see it worse sometimes, but it’s an average of around 50%. At the end of the day they do get the right people in place and there are people looking at it, but it all depends on when they make that shift.

And again an unexpected support cost often comes from- — again RPA is being sold as easy, simple, anyone can do it and there are components to that but the magic that it’s all going to be simply simple and one person can do any and all of it “Hey I don’t even need to switch to become a full-time developer, I can stay in accounting and just develop my own box on the side”.

It’s just not realistic so now often people are adding headcount they weren’t expecting and those headcount become more costly because PwC or another vendor is going to as soon as someone has RPA the cost go up. So that’s the experience.

Edward Brooks: There’s a whole study to be done in that, it’s a fascinating piece. We can often get involved and organization are just coming off the consulting, kind of thinking of support and they’re now looking to stand on their own two feet and they’re looking to up-skill their workforce.

It’s kind of a nervous time for them because they realize this business case we built has to kind of start happening and there’s a lot of uncertainty let alone the knowledge. But it is also fascinating because-

Edward Brooks: Absolutely and that kind of rolls into- -that kind of goes into knowing who to train and what to train in. I mean absolutely people need to be trained in this technology, this is not a Band-Aid solution, and this is something that people are going to do, that companies are going to use as another major enterprise software in their business.

So people need to understand it. We often push a lot “that everyone should go through the training”, you’ll be able to uncover who’s kind of got a knack for it, who doesn’t. We spend a lot more time on helping clients build out their business analysts to understand the process pipeline because at the end of the day you can do as much development as you want but if you don’t have the pipeline in your queue, you’re going to have a lot of developers that are just sitting there twiddling their thumbs.

So there’s a balance between getting the interest in all the different departments and getting things documented. The other thing, a lot of people don’t realize this, your processes are probably going to change because before, John in accounting had to run a report at 8:30 a.m. because he got in at 8 a.m.

After he redid that report it had to be checked at 9:00 a.m. by his boss because someone’s got to check the numbers, but neither of those two things need to happen because the bot won’t make a mistake, the bot can run it two o’clock in the morning.

So there is reengineering that is involved and then like I said there is a time and a place for off-shore but we have seen that more on the maintenance or very specific development where communication is being done by an on-shore on-site team.

I’ve had plenty of conversations with people that were outside the US, I told them the same thing. They are going “Oh hey we want one global business partner” and I’m like “Good luck with that but you’re going to be creating the same issue”.

So my statement was “Hey more than anything probably you should look for a partner that’s focused on your territory and wherever that back office sits”. I’ve seen a very successful off-shore/on-shore program with travelers actually because a lot of their back office is off-shore already.

So they’re sitting together and that’s a big plus but again, I’ve also had a lot of people say that “Hey, we’re struggling on the process of discovery and analysis” again wherever the main component is, needs to be done right alongside those team members.

Matt Gallo: Okay, there’s a whole subject in there totally and so in terms of preparing people, what’s the nature of events as the 90 days come to an end and then you have to kind of get the political buy-in or the budget buy-in and to keep going.

Matt Gallo: It’s funny within that 90 days our clients end up doubling down a lot of times to be honest because they see the success, they see where it’s going, big resource components of the RPA90X program is the six-month process pipeline, we share that very specifically.

It has the current process, the current amount of man hours that it takes, the complexity based on how many exceptions versus what it is their computer version that’s needed, how much time it is going to take to develop versus how much time it takes to do it manually and ultimately show them a development cost of what they’re going to save in the coming year, once you factor in licensing, software, maintenance and all those other things.

So yeah there’s a progression, general speaking the way it is, set it up, budget for it properly. A lot of times we actually see that being successfully done at the end of the year because- — or a part where they want to budget for next year.

So there are people that will start our RPA90X and then pause their program for a few months until they budget for the next year and then they’re like “All right now this is what we want to do”. So depending on the size, the reality is most companies end up saying “Hey we just want to AccelIrate RPA” and I try to get a dime every time someone uses the word “Accelirate” but I haven’t been successful in collecting.

Edward Brooks: Just to close off with these questions coming in, I mean a lot of the focus is on large organizations and how they use automation or RPA, what do you see at the smaller end for smaller service companies, consultancy businesses or accountancy businesses in the application of RPA for them?

Matt Gallo: Sorry are you saying the application for smaller companies using RPA?

Edward Brooks: Yep, exactly.

Matt Gallo: So there is a threshold, we typically see it right around a thousand employees, where anything less, I have seen- — it usually doesn’t end up going all right. One of the things I tell people-

Edward Brooks: Is this because of budget reasons or is it just the cost of bringing you in or is this because the technology doesn’t apply to them?

Matt Gallo: Technology will always apply but the biggest thing is- — my own personal statement is “If your goal is to get five bots in production then you’re just in the wrong area”. There’s just too many components to it to think about this as a five bot program.

It doesn’t mean you have to scale up to the thousands but at a minimum I tell people “Look if you can’t fathom budgeting like a half a million dollars for your first year, then it’s not really realistic to build out a program”.

That’s about that the magic number we see for a company to be successful with RPA. Year one costs are minimally going to be about half a million dollars when you factor everything in. That’s actually a conservative estimate.

So is it feasible? Yes, but half a million dollars to someone with 200 employees is a heck of a lot more than someone with 10,000 employees. So that investment doesn’t change whether you’re a 200 or 20,000 employees, it’s still a minimum.

Edward Brooks: But what we’re seeing is also a lot of opportunities are kind of at the desktop level it’s about a specific process level, but maybe it’s not full-blown RPA but certainly it’s huge, huge time savings or removal of transactional work.

A lot of that can be applied surprisingly cheaply, with limited knowledge, but it’s more of, I think perspective, the mindset and looking out for these opportunities, recognizing that they can be addressed, then you have to invest a fair bit of time to do the automation work.

I was going to say a number of things need to be started off internally. The first 90% of the journey is that easy but the last 10% it’s been pretty painful to actually go over the finishing line and so nothing happens fast but certainly the opportunities at desktop shouldn’t be ignored for smaller organizations

Matt Gallo: It will be, but here’s the interesting thing and this goes back to some of the difference in methodologies between some of the different vendors. I won’t get too much into that today but you should never be starting your program at the desktop level.

One of the biggest fears of RPA is shadow IT, I just saw it coming, if there are solutions suitable for all sizes and business needs, there are, absolutely, I’m not denying that, it’s just that when you actually factor in the cost of- — here’s the biggest challenge with getting started with RPA .

If you don’t have the right expertise in-house, if you don’t know the right- -you often don’t know the right questions to ask or how to properly evaluate people. I’ve seen a lot of times where they hire people just because their resume has everything they need on it.

You may be having this person lead you down a rabbit hole where you don’t know. If it’s done correctly, absolutely we’ve seen it be successful where it’s just a desktop automation and it does save a lot of time but generally speaking as a program, you don’t want something new- — you don’t want to add another software and another rendering, another thing to your business when it’s going to be such a very small component.

Typically best business practice is what we’ve seen.

Edward Brooks: Following on from that, just before we close off, do you see organizations taking responsibility for actually doing the implementation of the operation work or you just think there is an ongoing need for external support consultants to do that?

Matt Gallo: All consultants say “It depends”. What we see is once companies are bought into to RPA, their goal is to just go as fast as they can because they want to recognize those results. The faster they get automated the faster they start seeing those results.

So if they’re able to do it one term great, absolutely but a lot of times even if they build in that internal capability, we see that they are trying to just go faster. So the key thing, can you find a quality partner at a cost that allows you to recognize that ROI while you’re implementing it and then use that money.

The first thing that’s going to happen is the owner is going to invest half a million dollars of their own money, but in year 2, they’re probably going to have half a million dollars in savings. So now you’re playing with house money, once you are playing with house money, it becomes a lot easier to say “Hey instead of recognizing that half a million in savings, let’s double down and let’s get two million in savings and then let’s double down and get four million in savings”

And it often opens the purse a bit more.

Edward Brooks: Okay conscious that we’re at the end of the half hour, are there plans for RPA180X? Is there kind of an extended version, you think or RPA50X?

Matt Gallo: It all depends on where our client is, we do a lot of engagements where it’s very customized to clients, but what we did was say “Hey, where is the majority of our clients” and a lot of them don’t know what to ask for, they don’t even know what they’re looking for.

So we kind of packaged the key areas that avoid the shortcomings, make sure that the program is successful, gets it up and off the ground and in a reasonable amount of time. Absolutely depending on the client’s needs a lot of times as against them saying “Hey, we just need these 20 processes automated”.

And there are times where that happens, but in short the answer is, yes, but we like to package where we see the biggest needs.

Edward Brooks: Right. Matt, you’re doing fascinating stuff, really appreciate you taking the time to do this. Keep in touch and we’ll share your details with people online and this video will be online within the next day or so.

Matt Gallo: Sounds good, thank you.