Robotic Process Automation

Hello, and thank you for joining our fourth webinar in preparing for the future today series. My name is Chelsea Osborne and I will be today's moderator. I manage our central region commercial real estate deposit and payment solutions team. Our team members focus and understand the CRE business and work closely with our clients to help provide guidance on all things treasury management, including working capital, automation, and innovation.

Today, we are excited to share stories from three different clients within the commercial real estate industry who have adopted robotic process automation within their company. They will provide their thoughts on how RPA has made an impact on the day-to-day duties as well as provided efficiencies within their firm. Let's meet our panelists.

First, Cameron Cook, Assistant Controller at Real Capital Solutions. Cameron's focus is on two main items, corporate accounting and treasury operations. He is involved in the production of the monthly financial reporting, implementation of various systems, cash forecasting as it relates to dispositions and acquisitions, as well as overseeing the day-to-day treasury function. A big area of emphasis is on efforts to increase efficiencies and reduce costs across both the department and the company as a whole. In addition, Cameron holds an active CPA license within the state of Colorado.

Next is Danielle Enright, Tax Planning Manager at Vulcan. Danielle oversees the tax compliance for the investment management and professional sports team sectors within Vulcan. She is responsible for driving the automation initiative within the finance department and leads the Circle of Excellence team at Vulcan. Her past projects have focused on developing and implementing automation processes that save the tax department hundreds of hours annually.

Last, Denise DeRoss, Senior Vice President Corporate Strategy at Equity Lifestyle Properties. Denise has been with EOS for 26 years working in accounting, SB&A, treasury, and lease administration. For the past three years, Denise has been focused on improving the technology that is used at EOS.

Before we get started, I thought it would be helpful to provide a little background on RPA for those that aren't familiar. When I hear the word robotics, I can't help but think of R2D2 on Star Wars with the robotic voice and old technology. But that is the robot of yesteryear. Today, the word robot or bot has a different meaning.

Robotic process automation is the technology that emulates the actions of a person through interactions with systems in order to execute processes. RPA bots utilize the system just as a human would by following a strict script of steps in order to complete the task. By yes, no, and then-if questions, they can interpret and respond as well as interact with other systems in order to complete requests.

RPA bots can mimic most user actions, including logging in the system, moving data, completing forms, pulling information from documents, opening emails and reading attachments, and much more. In addition, bots can run 24 hours a day. With that overview in mind, let's get started.

As I just mentioned, RPA offers companies efficiencies and automation for day-to-day tasks. I would like to hear from each of our panelists on what made your companies interested in learning more about RPA. In addition, did any of you face any push back when thinking about RPA? And if so, what helped overcome those objections? Cameron, let's start with you.

Yeah, so our company, thankfully, is one that tends to embrace technology and creating efficiencies throughout our processes. And so last year, actually, at the US Bank Conference was kind of the first that we had really delved into RPA. And we were able to kind of bring back the concept and right when we got back shared it with a couple other team members.

And you could just see their eyes light up. And we kind of started to discuss different applications that we might be able to use it for and kind of started to do a lot of research there internally. And that kind of just brought us to where we are today.

We did-- I don't necessarily think that we had push back per say, but I think there was some nerves within some people thinking, OK, is RPA going to take my job? Is it here to replace me? And so I think that was the biggest hurdle that we've had so far is just trying to educate people about what RPA is and how it's not necessarily to replace someone but it's to add those hours back to their day so they can do the more valuable tasks that we have throughout our organization.

Yeah, that's great. And we didn't pay him for the shameless plug. That was voluntary.

Yes, that was.

Denise, what about you?

Yeah, so I think for us about three years ago I was talking to our executive team about what we needed to focus on for the company. And we really believed at that time that technology and process automation were really important things for us to focus on, that we really needed to get our technology moving forward. And we needed to improve what our employees were able to do, right? Not having them focus on manual processes but have them focus on really value-add pieces of their job.

And so we kind of came up to this idea as a group. And therefore, I didn't get any push back. There was a really big embrace from the entire executive team to move forward with RPA. They loved the idea and were behind it the entire time.

That's awesome. Great news. What about you, Danielle?

So we were a little different in that we were actually approached by one of our service providers, BWC, for an RPA solution that could help us on the tax side save a bunch of time. And that kind of got us thinking about what other processes we could automate. And we kind of had the same push back like Cameron mentioned with people just really needing to educate them and let them know we're not trying to replace your job with a bot. We're really just trying to get you out of doing the mundane tasks. And you can actually spend time doing value-added processes or researching or being more analytical, which is hopefully the more exciting part of people's jobs rather than just doing the mundane, day-to-day exercises.

Yeah, absolutely. I think you bring up a really good point there. Well, thanks for that insight on a little bit about how your companies learned about RPA. I think that's really helpful.

And as we know, bots can provide benefits to many departments and groups within each of your firms. But I think it'd be really interesting to hear on how RPA has been utilized within treasury first. So Cameron, we'll let you start. And if you could share a few use cases about how your team has adapted bots in treasury AP and AR, that would be great.

Yeah, absolutely. So one of the first items, I guess, that we have kind of adopted is on the AR side of things. And it has to do with payment remittance.

Every month, we receive a hundred plus payments electronically. And with those payments we're actually emailed a remittance file for each individual one. And so previously, somebody is getting the emails, they're printing out the back up, and then they're having to go into the accounting system to actually record these payments.

Now what we've done utilizing a bot is the bot actually receives the email, compiles the support, extracts all the data from the email, puts it in an Excel format, and then actually imports that information into the accounting system. So not only does it kind of take away from the manual entry on our part, it gives us that time back to actually review and kind of high level make sure everything looks correct. It eliminates errors because the payment application is driven by the remittance form.

So there's never a chance of us accidentally keying in the payment to the wrong charges, things like that. Which, when you kind of look and evaluate how much time is spent on the process, I think, at least for us, we tend to leave out the time spent after the fact of researching issues that have come up because of human errors, whether that's hard keying a number incorrectly or applying it to the wrong charge. So that was kind of one that we kind of estimated that's probably about a five hour savings per month. So while it's not a ton of time, now that's just one little piece that starts to add up.

And I think probably one of our bigger ones, actually, which was kind of the first one that we tackled was the bank reconciliation process. And so there's kind of multiple aspects to that, one of which is we use a bot to actually log in to each of our banks and download the bank statements on a monthly basis. And so for us, that's 15 to 20 different banks. And that's 300 plus accounts.

So while it's simple in nature, it becomes a little bit of a time-consuming task. And so now the bot just logs in, downloads it, and actually saves each of the statements in a very specific format. It's unique for each one of the accounts, so it makes it really easy to pull up that particular statement and know exactly what it is that you're looking at.

And then we kind of took that even farther within the reconciliation process where we're able to-- well, let me take a step back. So we actually-- we reconcile all of our accounts through our accounting system. And so within that process there's a lot of manual clicks where you have to log into the accounting system, open up each account individually, enter information from the statement, clear all your transactions, print out your report.

And so that time starts to add up pretty quickly. For us, we kind of estimate that that's about 70 hours a month that we spend reconciling bank accounts. And so now that we have RPA involved and we have a bot that can help us with this, in addition to I guess some other automation tools that we have throughout the process, but we're essentially eliminating somebody from having to go through the 35 clicks it takes to actually produce this report.

And instead of them spending time doing that, they're now able to spend time reviewing the actual reconciliation, making sure that everything ties out correctly, and bringing up any issues, flagging any items that actually need to be looked at. So yeah, it saves a ton of time for us. But again, it also tends to reduce human error and allows for the more valuable work to be done. I think those are probably two of the treasury-related tasks that we've started to use RPA for.

That's great. And I think you said it a couple of times, eliminate errors. Like, that's a huge theme I think we'll hear more about today as we talk.

But a bot doesn't make human mistakes, transposing a number or entering a GL entry wrong, those kinds of things, which is huge. Denise, I think you guys also utilize RPA for bank reconciliation. So maybe can you share some other uses that your company is using bots for within treasury AR, AP, and accounting?

Sure, absolutely. So our first bots that we put in place actually really focused on our accounts payable area. The first bot we did was for our new vendor set-up process.

So we had an AP person who spent their entire day setting up new vendors. We had a 72-hour turnaround for new vendors being set up, which created quite a bit of angst at our community level because it took that long for them to get a vendor set up. And we had a lot of human error where things were mistyped, addresses were incorrect.

And in addition to that, we actually had some separation of duty issues because you can't have the same people setting up vendors that are inputting invoices. And so we were really limited to having this one person do all of that activity. And what we did is we went forward with a bot where we created a form in SharePoint where our operations people could do the new vendor request. The bot would read that request from SharePoint, put it into a table, and then start making decisions on that.

It would determine from our ERP whether or not that vendor already existed. And if it did, it would send a communication back to the requester with the vendor number telling them that it already existed. And if it didn't, it would go through and actually it clicked through all the different fields in our ERP system, sets up the vendor. And at the end of that, it sends an email to the requester telling them the vendor was set up and providing them the vendor number.

And that process runs every hour throughout the day. So now we've turned a 72-hour SLA into a one-hour SLA, which the operations team loves in comparison to the 72 hours. It's probably been our most successful bot as far as people-- the effect across the company, how many people are affected by it and has the most obvious benefit right out of the gate. No question about it.

We also did a couple of other bots related to AP. One was invoice coding. So all of our invoices would come into our AP system without any coding on them. We have a fairly old AP system, so it doesn't really have, like, previously used account or anything like that.

And so we were able to put this bot in place that grabbed the history of our vendors and where we coded payments to on a location and vendor basis. And then it was inserted into the workflow so that every invoice that comes in goes to a specific queue for the bot. The bot looks at that queue and starts determining whether or not it has an answer for that location for that vendor.

Does it know historically where we coded it to? And if so, it codes it. And if not, it just leaves it blank. And in either case, it pushes it forward in the workflow and allows the operations person to review that coding and change it if they don't like it, they don't think it's correct, or enter it if the bot wasn't able to find the coding.

Pretty much right out of the door, we were able to code about 70% of our invoices using the bot so that that was pre-coded for our operations team and they just needed to double check that. So we process about 250,000 invoices a year. So that is a lot of invoices that get pre-coded, which is a huge benefit to us for sure.

Yeah, that's great. That's a huge number. And Denise, you shared a little bit with me when we were chatting earlier about maybe some training opportunities to help make that bot more successful. Can you maybe share with the team what you're seeing kind of as next steps or ways to make the bot more effective on that front?

Yeah. One of the things with the bot, it's looking at history. And if the history happens to be wrong, then the bot is wrong.

And so then it doesn't really fix either having the operations team have to fix the coding, or if they don't fix it, having our accounting team have to do a journal entry re-class on the back end. So what we want to do is have our accounting team be able to go back into the tables for the bot and say, this is what the actual right coding is for this particular location and vendor combination so that we could get much higher than 70%. We think that we could get something probably closer to 90% of our invoices coded if we undertake that process.

Yeah, that's great. So it's almost like it's an always moving, always growing, always improving process. The bot's great out of the gate, but there's always ways to make it better as well. So that's something-- I think that's important to keep in mind.

Absolutely.

Thanks, Denise. All right, Danielle. What about you at Vulcan?

So I would say our most successful bots have been more on the AP side of things. So the first one that we implemented seems like a super simple process. It's just loading foreign currency data into our ERP system.

So before we had the bot, someone would go in, manually look up about 20 different foreign currencies, and then one-by-one and put them into our systems. And as you can imagine, pretty monotonous, boring task. A lot of room for error. You could hand key in something wrong. And then on top of that, someone else would have to come in and review just because there was a lot of room for error.

And so one of the first bots that we did would actually go onto the website, look up every single foreign currency, input it into an Excel file, and then plug that data into our ERP system. And so this is done every single month. I think we just have it scheduled to run on its own so no one has to go in and trigger it or anything like that.

And we estimated that it took someone about two hours every single month. So not a super heavy lift, but again, annually that's 24 hours. That's three full days that someone's spending inputting foreign currency information into our ERP system.

And so while someone still comes in and reviews everything to make sure everything's correct, we've completely eliminated the having to pull the information down. And I think this was a really good example of a pretty easy quick success to get people on board. And it's easy to prove the benefit and eliminate someone from having to do this every single month.

And then another one that we do is a sales and use tax bot. And so we're located in the state of Washington, so we something called a sales and use tax. And every single month, someone from our AP team would go in, pull about five different reports, aggregate all that information on every single payment made, which some months it could be thousands, some months it could be a hundred, it just kind of depends.

And then they would-- we're supposed to take the initial review of that information. But one thing that we found is since they've spent so much time, like probably four to five hours just pulling the data, aggregating it and organizing it formatting it, that they were really struggling with kind of taking a step back and being analytical and reviewing the information. And so then our tax team would have to come in and review everything and then decide if we needed to pay additional taxes on the payments.

And so now we have a bot. It's kind of a two-step process that we use. So we initially scheduled the bot to run the first of every month. And it pulls all the payments, aggregates all the data, puts it into a single Excel file that's easy to read.

And then we also created a couple of macros that kind of go through a variety of rules. So say if it's hitting a certain account, we know it's not-- we don't have to pay tax on it so we don't need to worry about it. So it's automatically eliminating those payments.

So now our AP person can actually come in and be analytical and review. And it has a smaller set of data to review. And they can spend the time, instead of just pulling the information together, actually reviewing the information and thinking through if it makes sense that we need to be paying tax on this. And then they give that information to our tax team, who then we file the tax returns.

So a pretty monotonous process again, but it really allows our AP team to be like the first level of review rather than our tax team having to do everything. And it saves on average about ten hours a month, so pretty big bang for our buck. But I mean, it did-- it was a more complex bot and it took a little bit longer to develop.

But I think we're going-- we run this by every month. We're going to be running it as long as we're in existence because we're always going to have to be applicable to the state and local tax. So I think over time it's really going to have a pretty good ROI.

Yeah, absolutely. I think your two examples, and the examples that both Cameron and Denise shared, you can see that a bot can run a very complex process or very easy process. But there's value in both. So I think that that's important to highlight as well.

So clearly we know that a bot can run a lot of processes within treasury. We ran through a few examples as a team. But Danielle can you share a bit about how you have been able to utilize a bot for a one-time use?

So we had a pretty unique situation because of COVID. We own a handful of theaters and museums that people could have annual memberships to. And obviously for the majority of 2020 they were not able to benefit and use the facilities. And so we decided to issue refunds for the annual memberships that they paid.

And so we had thousands of memberships that we needed to refund. And the process would have been to go into our ERP system, set up a one-time call it supplier just to issue the payments. So someone would have to manually enter all of that information one by one by one, and then someone else, a manager, would have to come in and review and make sure that everything's in there correctly.

And so we figured this would be a really great case for a bot. And even though it is one times, we thought that it would be worth it because we estimated on the low end that it would take someone about a month of working just on this just to input all of the information. Just because-- I mean, our system isn't the best. It's kind of slow. So having a bot do it is a lot quicker.

So we actually just pulled all the information from Excel, and the bot would just pull that and put it into our system. So someone still had to come in and review it, but again, eliminating the errors of someone hand keying in something wrong. Also, just making someone spend that much time inputting information one by one by one into our system, I don't think anyone wanted to do that. And so ARP was a great solution for that in that we could eliminate that entire step and someone just had to review the information and approve the payments.

Yeah, that's great. I think a lot of folks probably think of using a bot, it needs to be something that's repetitive, that's long term, that you're going to use a lot within your company. Which is very important, and you have to obviously offset the cost in the development and all of those things that are involved with the bot. But clearly this is a really great example of you guys knew that the impact that it would have and the time saved and it made it worthwhile to go down the road. So thanks so much for sharing that.

So we clearly talked through treasury and we covered those. And then Danielle, thanks for that one-time use. So I thought it'd be helpful for you guys to share a few use cases outside treasury AP, AR. So Denise, let's start with you at ELS.

Great. Yeah, our head of marketing was very interested in the RPA and how that might work. And so she actually brought us a couple of use cases for RPA that I thought were really interesting because they were outside of the realm of how I was potentially thinking about it. And the first one she brought us was for Google reviews.

So we own and operate manufactured housing and RV communities all throughout the country. And so people leave us reviews on Google. And if they put some text into the review, they give us a star rating and they put some text in there, we have an employee who can go in there, read the text, and then respond accordingly to the text that's provided.

But a lot of the Google reviews are just star ratings. They're not-- they don't actually contain any text. So there's nothing that a human can really do beyond say, hey, we got a four star, thanks for the rating sort of thing. And we couldn't even get to those because we get so many reviews that our one person couldn't even consider getting to the ones without text.

So we had the bot go and look at all of those reviews that only had star ratings. We provided the bot with a handful of options for each level of star rating so that the bot wasn't answering the exact same way every single time. And now every evening the bot goes in, finds those reviews that don't have any text, puts the text that we've provided to the bot in there, and responds to those reviews.

And while it's great that we're able to respond to those reviews, it also changes how Google rates you as a company. If you're responding to more of your reviews, you get a higher rating. Google sees you as a better company. So that's just a side benefit of being able to respond to all of those reviews.

And then the other thing we did was we're in process of getting this one going, which is if Expedia and Booking.com, we list our cabins and homes that we have at our RV properties for reservations on those websites. And in the past, we've had an issue where some of our properties were not showing up as being available for-- like we had one that wasn't available all summer long. And that's clearly a problem. It's hard to get revenue through those sites if your sites aren't available. So what we're trying to do and trying to get in place is have the bot compare what we sent to Booking and Expedia to what they have on their side and validate that yes, I can make a reservation for this property for this time period as expected based on the availability that we know is correct on our side. So we're hoping that once we get this bot in place that we would identify any time there was some sort of data issue on the Expedia or Booking side that prevents our customers from being able to book our sites.

That's uh, yeah I can see that as being really beneficial, especially for your properties that aren't showing up on there all summer, which I'm guessing is probably the busy season, right, for some of those sites. So really important. That's great. That's really creative use of a bot. Cameron, what about at RCS?

Yeah, so kind of going back almost to what both Denise and Danielle have talked about on the AP side, we've kind of taken that approach on the accounting side just on the administrative setup. Not only of vendors, but every time we acquire a new property we can now utilize the bot to actually set that up within the accounting system. And same thing with bank accounts.

Every time we open a new bank account, we have a set up form that eliminates the human error part of it again. And the bot can just automatically take that information and import it into the accounting system. And yeah, I mean, that one, while they don't happen as frequently as say, like setting up a new vendor all the time, I think there's still a lot of value there. And that it does reduce the chance for error.

Another probably another one that we've started taking on is transferring files around. So every day we receive a file via multiple SFTP sites. And we're able to utilize the bot to actually go log in to the site, grab those files, and then we do different things with certain files.

But one of them, for example, we actually take it and we import it into the accounting system. So the bot then basically does this whole process for us. It goes in when there's-- well, sorry, let me take a step back.

We receive an email when there's a file that's ready to download. The bot can then read that email, log into the site, download the file, save it to our network, and then import it into the accounting system. So definitely eliminates that human element there, that mundane task of moving a file around. And then allows for somebody to actually review the information that was on the file instead of spending all the time just moving it around. So I think that's probably a second application.

And then the third is just kind of extracting data, whether it's from emails or from websites or as we mentioned about the foreign currency. Maybe we're pulling LIBOR rates or things like that. It just makes it a lot easier and gives people some quality of life back in their job, I think.

Yeah, that's huge though, I think, especially kind of in this COVID environment we're in where everyone's always accessible. To have a bot do some of those manual, kind of tedious tasks that some of us may do at nine o'clock at night, obviously, like you said, brings back some quality of life that I'm sure your employees appreciate.

Yeah.

Good. Danielle, what about at Vulcan? Any uses outside treasury?

Yeah, I would say we've had a handful. One really good example is our cost control team. So we obviously develop a decent amount of real estate.

And so daily they're kind of obviously tracking the costs of what's going into different projects that we have going on. And before we had a bot, they would go in and manually pull down the cost reports for every single project that we're working on, which can be five to ten different projects at a time. And now we have a bot that just goes in, pulls the data from every single project, and emails it to everyone on the team.

So pretty basic task again, but someone would spend probably almost an hour every single day. And I think even before this they were only doing it weekly or monthly just because it is kind of a pain and not a high priority. But it's nice to have the information. And so now we just have the bot run every single day so they can always have up-to-date information on the projects.

Another good one that we actually worked with on with our HR department is kind of going through and seeing which employees. If someone's left or we have a new employee, like making sure they're completely set up in our system and that their managers are fully listed. And we have little trees that you can go into, like our employee website, and it shows who everyone works with.

And so now we have a bot that makes sure that all that information is up to date. Because before, someone would have to manually go in and update everything. And so again, not a super high priority, but it's nice having everything updated. And the bot can run-- I think it runs once a week now just to make sure that all of the information is up to date.

And then we've had a couple of other ones where if we had a new system implementation, kind of like Cameron was saying, it's just pulling data and dragging it into our new system. Again, another one-time case, but it was pulling records back to 2010 and uploading all of the information. This was for our investment management side.

But uploading all of information for all of our investments and making sure it's loading correctly. So a pretty tedious task, but something that was really easy to do and accomplish with a bot. And it was a great case, even though it was a one-time process.

Yeah, those are all really good use cases that you guys have shared. I think the impact within treasury and accounting, obviously there's a lot of great examples there as well as other departments. I mean, a bot can really touch almost all facets of your company.

And I think, again, it's important. Like, it doesn't have to be a full-fledged process. It can be just extracting data from websites or systems.

One thing, the US Bank is actually starting to use bots as well. And one of those areas that we're starting to use bots in is our account services spreadsheets that we put together for our client reviews that we have. And as you guys are aware, there are a lot of line items on an analysis statement. And there are a lot of accounts.

And so in the past, our account assistant relationship managers actually manually put those spreadsheets together based on the analysis statement. So they would print out the analysis statement and fill in the spreadsheet. And we have started to utilize a bot for that process.

And so now the bot pulls in the analysis statement and then reads the analysis statement and put that into an Excel format that's user-friendly. And then our team can go in, and speaking of value add, highlight where we're missing services. And instead of spending their time trying to figure out that one account they're off because they manually input it, it's actually able to go in and highlight like the things that really are important, like you're missing fraud prevention on this one account or whatnot. So obviously there's a lot of benefit there. And again, it goes back to having more value add tasks versus those tedious, monotonous tasks that I think everyone, like we shared, appreciates getting off their plate. So thanks again for sharing those great use cases.

Moving on. With all those examples we talked about in mind, I thought it would be helpful to hear from each of you on the steps your companies take to determine what processes would really benefit from the use of the bot. So Danielle, we'll start with you on this one.

Yeah, so we actually have. We call it a Center of Excellence team. And it's a representative from pretty much every single department within finance.

And we started this with going through a two-day training. So we actually are the ones developing the bots. But part of our process now is we're supposed to be talking with our groups, our teams, to see what processes people are doing. If they think they have a good case, then it's the person, the representative's, job to bring it to our team. And we meet every single month and just kind of run through the process.

I think one thing that's been really helpful for us is that you kind of flowchart the process and see, like, step by step by step because that is how the bot is built. So one step that you might forget or think doesn't really matter obviously matters when you're developing the bot. And so we kind of run through what the process is and then as a team we decide if we think this will be a good candidate for RPA.

And we look at how many hours the process is taking right now, how many employees, if it's multiple people involved. And then we try to estimate how long it takes to develop the bot. That's kind of the trickier part, just because there's a ton of testing involved and you just never really know what issues can come up.

Yeah, so we just make sure that we get the buy off from every single person in our team. And then it's on that person. So if it was a tax bot, then it would be on me to develop the bot, which is kind of a unique, I think, citizen-led approach compared to other people.

Yeah, that's really interesting that you have your team that kind of determines from each department. So it's great. Cameron, what about at RCS?

Yeah, so selfishly, and I guess luckily, we kind of started off as almost like a proof of concept of the bot within our company. We started looking at tasks within our department on the corporate and treasury side. So we were able to kind of identify what we thought would be really good candidates to start with.

But now that we're starting to kind of branch out, and I think others are able to see what these benefits are, there's kind of a lot of ideas that are floating around out there of what might be a good candidate for a bot to look at. And so what we've tried to go back and do, similar to what Danielle mentioned, was actually documenting out what your current process is. So we're pretty good now at making flow charts to identify every step within there.

And I think what that does is not only does it allow you to look at what you can use a bot to possibly do, but it also allows you to just identify unnecessary steps or ones that you can tweak that don't necessarily even involve a bot. But so in going through that process then, we're able to kind of put together an estimate of what the current time hours are that go into that process. And then we also can look at what would be the perceived benefit of this. Is it hours that you're getting back? Is it a reduction in cost from somewhere else?

And knowing kind of both the benefit that you can achieve and the effort that it takes to go into it, we've started to develop a scorecard. And our business application manager, he's kind of the one who's led that of putting together this matrix of what are the most high-impact tasks that we can look at using a bot for. And that way, if we have a really high impact and a really low effort that goes into doing it, well, that's probably going to be our number one candidate that we look at now versus something that's low benefit but takes a lot of time to do, probably going to put that one lower on our priority. So that's kind of where we're at today.

And I mentioned going to the US Bank conference last year, well, I guess two years ago now. But towards the end of '19 this is still relatively new for us. So this is what we're-- this is how we're approaching it today, but not to say that that might not change in the future.

No, I think that definitely makes sense. It's really important to probably prioritize what makes the most sense and what's going to have the biggest bang, right? Bang for your buck.

Yeah.

Denise, what about at ELS?

Yeah, so the way we do it at ELS is we just have anybody who has a manual process that they would like to have automated, they bring it to me and my team. And then we take a look at that and determine-- kind of help that person determine whether or not a bot is even the right answer. Because it might not be, right? There are a bunch of other solutions that our IT team can work with them on, try to vet out that process, make sure we understand what the steps are, who's involved, how much time does it take in the current process. And then how much time would it take for a bot to be created and put into place. So we do that. We have everything kind of coming through that group so that we can kind of redirect based on what the particular need is for the solution to that particular request.

Yeah, that makes a lot of sense. Thanks, all of you, for sharing your thoughts there. I think it's really important for everyone to understand what goes into really determining what could benefit from a bot or what processes could benefit from a bot.

Cameron, you touched on this a little bit. But I think it'd be great to hear if your companies have really been able to determine a value or a return on investment for RPA. And if so, how you've been able to share that with leadership at your companies. Cameron, we'll start with you.

Yeah, we have definitely tried to quantify what the return on investment is so far, which can prove a little challenging because one of the things that we look at is scalability. So while something might say a number of hours today, well, if we multiply that tenfold because we're expected to grow by that much, well, now we've saved by having not to hire another four or five bodies to do something. We kind of look at the ROI as a, I guess, multi-tiered approach here.

One of the big ones, as I was talking about, the bank recs and the number of hours. We're able to extract that number of hours on a monthly basis and a total dollar amount, give or take because the labor cost is different for everyone, I guess. But yeah. So we've been able to do that so far, and within year one, I think we've already more than recovered what our initial investment has been so far. And that's just-- I say year one. Honestly, it's kind of like a six-month thing where-- when we actually started to deploy the bot. And yeah, already recovered our costs and just looking at how we can continue to grow and what that return really equates to.

Yeah. That's great. Thanks, Cameron. Denise, what about at ELS?

We've actually focused more-- we have focused on how many hours of work has been saved, but we've actually focused that in a way of improving the satisfaction that an employee has in their job and-- so especially with those bot that are in our AP departments, in our accounting departments, we had a ton of really manual processes there. And we had people who spent their days doing those very manual processes, and so that doesn't give them a lot of job satisfaction. It can make it very difficult to keep those employees in place. They leave because they don't want to be doing the manual bank rec process.

And so we've really focused on that piece, everywhere from accounts payable and accountants being more satisfied. And also kind of the recipients of those bots, our operations team, improved satisfaction with our whole process because they're getting better responses and more information when there is an issue. So while we do pay attention to how many hours are saved, I think this idea of more satisfaction with the job is really where we focus more.

Yeah. That's great. And I think we heard kind of the same sentiment from Cameron earlier just talking about quality of life and getting those tedious, monotonous tasks off everyone's desk, which is great. Danielle, what about at Vulcan?

So I would say we're kind of a combination of both. We have a tracker that shows every single bot, where it is in the process. And if it's been completed, we include how many hours it spent to develop, how many hours it's saving, and we use a blended rate to estimate labor cost savings. But then we also look at what the bot is doing. Is it letting someone spend more time being more analytical rather than being in the details or pulling down reports and just aggregating data?

The one-time bot we talked about earlier, I think anyone who would have been asked to do that would have not been very happy to spend that much time doing the same task over and over and over again. And so we also take that into consideration when our team is deciding if it's a process that we should use RPA for because we have to look at both, and I think keeping both in mind is important. While, obviously, dollars saved is important and we want to save more money than what we're spending for our licenses for RPA, I think job satisfaction is also a really key component of it too and making sure that people are happy and they're not spending their time doing mundane, redundant tasks.

Yeah. That definitely makes sense. And I think the sheer hours saved speaks volumes of really how beneficial RPA can be. Not only-- like, Cameron, you put a dollar amount or you're trying to put a dollar amount on the hours and what the labor costs would be. But the number of hours just saved and your team doing other tasks is really huge.

Before we move on to our last item, I thought it would be great to hear from each of you. Danielle, you touched on how Vulcan has developed your bot in-house. Can you elaborate a little bit more on that process? And then we'll let Danielle-- or, excuse me, Denise and Cameron speak about how they've been able to engage bots.

Yeah. So like I said before, RPA was kind of brought to us by PwC, which is one of our service providers. And from that, we kind of looked at and met with each department about if we thought processes could be automated. Obviously, I think every department has room for automation, but we really wanted to create a proof of concept and kind of show that it's worth spending the money of working with PwC.

And so once we did that, we went through every single department within finance and asked for a volunteer. Some people were voluntold, but everyone-- every department got a representative, which I think is really important because we have a lot of processes that involve more than just one department. So I think having every stakeholder in the room is really key to the success of it.

And so we went through a two-day I call it boot camp with PwC learning the ins and outs of UiPath, which is the system that we used. And then part of that boot camp is every person was asked to bring two to three different cases that they can work on. So we went through a bunch of learning cases and then also started developing and working on our processes.

And then we had someone from PwC who-- this was pre-COVID times-- would come into our office. And they were there for, I think, about three weeks and just had open office hours, which was really beneficial because while UiPath is pretty easy to use, I think there's a lot of nuances and one-off things that you just don't-- you're not expecting to come up and you need the help with. So having someone there to physically help you is really nice.

And from that, we just created-- we call it our center of excellence team, like I said before, and we just meet monthly. And everyone is kind of on their own with developing their bots. We talk about the bots. We, obviously, approve it as a team. But if it was a process that was related to tax, then it would kind of just be on me when I have time to be developing the bots. So kind of a unique approach, but I think it's really nice to have us being the ones doing it just because we know our systems the best and the ins and outs. And whatever troubles and issues that we have with the systems, we are the people who know those rather than bringing an outside person in.

And I think at Vulcan, because we have so many different segments and so many different lines of business, every person really is an expert in their area, whether it's real estate, finance, or investment management or AP or tax. Every person really knows the ins and outs of their systems, and every department and group uses different systems. It's not everyone using the same thing. So I think having our team and having the representatives from every department has been really beneficial just making sure we're getting the best use cases and the best return for our bots.

Yeah. That definitely makes sense. Denise, what about at ELS? Did you guys utilize a consulting firm, or did you develop your bot in-house?

No. We decided to go with a third-party consulting firm. We have a lot of other automation processes that-- manual process we're trying to automate and new technology that we're trying to put in place. And we thought that we could accomplish more if we went to a third-party consultant. And then at some point in the future, the idea would be that we could bring that in-house when we have more capacity in our IT and accounting departments in particular to take on that internally and not have that third party involved.

Yeah. That makes sense. Cameron, what about at your company?

Yeah. We kind of have a hybrid approach there, a little bit of both. So we were working on another automation project kind of when RPA came to light for us. And with this other automation project we were working on, we actually had a third-party consultant that was helping us out with it. And they just happened to bring up RPA as a potential solution for that project. And so we thought, well, shoot, maybe this will be a good idea to kind of work with them and understand RPA and almost just use them as a resource to do some research and things like that. So we did.

We kind of had a month-long engagement with them where we went out and we used some of the free demos and started to review the different vendors that are out there alongside of the consulting group so they can show us what might be more applicable to our business and how we want to handle it. And after that month, we had thought through some of the best practices and things like that. And we kind of are just handling the bot development internally now.

I think that Denise brings up a good point about is it more efficient with all the different tasks you have out there to bring it in-house, to have third party help with it? And now I think where we're leaning is still to keep that in-house and maybe even look to bring in another resource to help out with that automation and development.

That's great. You each kind of have different stories, which is really important, that it isn't a one size fits all kind of engagement. So thanks for sharing that.

All right. Last question. In looking back at the implementation of RPA from start to finish as well as all the current usage you guys currently have with RPA, what lessons has your company learned that would be beneficial for the audience to have in the back of their minds should they want to implement RPA? Danielle, we'll start with you first.

Yeah. I think the biggest lesson I learned through leading this project is making sure you have all the stakeholders in the room. So we have, obviously, our center of excellence team, which is finance, but there's a ton of other groups. And I think every other group might have better solutions than RPA. And so we actually had one case where we went down, thought RPA was going to be the best solution to update a bunch of vendor information within our system. We had someone developing it. They had spent probably 40 hours working on this development of this bot.

And I kind of had the idea maybe we need to bring in IT just to make sure because we were having some issues with it working with our system. And they posed the question of, oh, why don't you guys just write a script for this? That would be way easier. And I'm a tax person. I'm not IT. I didn't know that that was an option.

And so that was kind of a big lesson learned for me in the sense of making sure you have all the stakeholders in the room because sometimes RPA isn't always the best solution. And maybe your system has automation built in with it, or your IT department has a solution that is easier to develop or create than creating a bot. And so I think that was a really big lesson learned, and I felt really bad for the person that had spent all this time. But I think it was a really good learning lesson at the same time of just making sure you're really getting buyoff from everyone.

And I think the other thing is just kind of like we've talked about before you go down the process of developing a bot is really flow charting it out and thinking through step by step what your process is because-- I think Cameron touched on this, but maybe you can create efficiencies within your process that don't even involve RPA and just making the process more efficient in general. And sometimes I think people don't really think through the process. It's like, oh, that's just how we've done it in the past, so that's just how we're going to keep doing it. And so I think there's always room for improvement. And when you're looking through potentially automating a process, flow charting it out and thinking step by step by step and what's the best solution is really important.

Yeah. Those are all really, really important things to consider. Denise, what about at ELS from your experience?

Yeah. I think ELS learned a lot of the same lessons that Danielle was bringing up, making sure everyone's in the room. And I think just building upon that, also making sure that the departments that are affected are really committed to it, right? This isn't a drop and run sort of thing where you say, oh, I want this, here you go, and then you don't have to participate anymore. They really do need to participate in the flow charting and the understanding of it and making sure that it works. And some bots have-- like our AP coding, they have some backend improvement processes that need to be done so that the bot gets better and better. So you need that engagement.

And then you definitely need to let your IT department know for sure what you're doing. When I went to the IT department and said, yeah, you need to set up these users, they're going to log into our system, and by the way, this user is actually a bot, that tends to send your IT department's heart fluttering a bit that you're letting a bot hit one of your main systems. And so we had to stop our process and then kind of backtrack and walk through all the steps with IT and make sure they understood what we were doing and let them get comfortable with the whole process, and then we could get our IDs and everything set up. So that did set us back because we didn't involve them right from the start.

And I think the other thing is, along the lines with what Danielle said, that RPA isn't always the answer. And I think that that's why it's really important-- I have our head IT developer in every discussion so that he can say, yeah, I could do that. I don't have to use RPA for that. We're going to go write a script to do that. There's a much better way to do it. And that we can just kind of knock that out from the beginning versus get further down the path and then realize that we took the hard way instead of the easy way. And so it's great to have all of those people in the room for sure.

Yeah. That's really helpful. Thanks, Denise. Cameron.

Yeah. I think they kind of covered it so far. I think the only thing that I'd like to add is evaluating what you've actually put in place. Don't just implement a bot, roll it out, and then never think about it again. I think taking the time to look back and see is the bot accomplishing what we really set out for it to do, and if not, then maybe it's time to rethink the process or rethink how the automation is going. I know that we've tried to spend some time doing that the past month or so of going back and fine-tuning things. Sometimes it's kind of duplicating efforts, but if you can take the time to do that now, I think it sets yourself up for greater success down the road. And that's probably the only thing I'd add to what Danielle and Denise had already brought up.

Well, thanks for sharing. I think that's all really valuable feedback and important considerations for anyone who is thinking about RPA to keep in the back of their mind. So I'll conclude by saying thank you so much, Danielle, Denise, Cameron, for your time today as a panelist on our RPA webinar. The insights you have shared during our time together have been very beneficial to our audience and will provide a really solid background on what goes into implementing RPA as well as ways to adopt the same technology within their company.

This completes webinar four in our four-part series on preparing for the future today. Shortly you will receive a survey on the webinar, and we really value your back. The feedback you provide really helps us determine our future content, so please provide your thoughts. Thanks for joining us, and be on the lookout for additional information on our next webinar series. Stay safe, healthy, and we wish you a very successful 2021. Thank you.