This post includes the full Transcript of from November 2018 – Ashling Partners Webinar Leveraging RPA to develop a ‘Best in Class Vendor Management’ Capability on November 23, 2018 and the YouTube video of the entire educational webinar: Leveraging RPA to develop a ‘Best in Class Vendor Management’ Capability
Ashling Partners and UiPath discuss traditional challenges in the Vendor Management and Procure-to-Pay processes, and how to leverage Robotic Process Automation (RPA) to create opportunities in a broader Source-to-Pay process.
Introduction: Transcript from original webinar recording
All right, thank you everyone for participating in our webinar today, Best in Class Vendor Management, addressing the challenges within the source to pay process with RPA. We’re going to kick it off here with a quick introduction. So with me is Marshall Sied, one of the co-managing directors of Ashling Partners. My name is Don Sweeney. And we’re also very excited to have Chet Chambers on the line, Vice President and Chief Evangelist in the Americas of UiPath. Chet’s going to go through some lessons learned that he’s seen with RPA, we’re going to start with that deep dive. And then we’re going to bring it back out a little more broad, and talk about some best practices and some KPIs within the procure to pay process.
So quick agenda here. We have a leader’s view on scaling RPA, as we talked about, we are very excited to have Chet. We’re then going to go into those challenges in the procure to pay process, talk about more best in class, the future of procure to pay and vendor management, the impact of Intelligent Automation in that kind of broader source to pay, even extending procure to pay to source to pay, and then really ending with how to make this sustainable. Before I turn it over to Chet, I did want to briefly mention a couple items here. With RPA, the robotic process automation, we’re really starting to see the value of the convergence of things like cloud, digital, artificial intelligence, all of these now starting to impact vendor management, and starting to impact the expectation of vendor management, around need for more speed, more flexibility, more precision, and really trying to get things done better, faster, cheaper than we’ve done them before. And so now going forward, we’re going to turn that over to Chet.
Thank you very much, and appreciate you having me. So my name is Chet chambers. I’m the Chief RPA Evangelist for UiPath. For the Americas. I’ve been doing RPA for almost a decade, I’m what you would consider an early adopter. I saw early on that there was nothing that was going to change the market the way that RPA is. So in those early days, it consisted of screen scraping, it consisted of being able to manipulate small pieces of data, and boy has it really changed over these last 10 years. So I was working for a company called VHA that was using automation anywhere as an RPA tool. That was the tool that got me started, I was fortunate enough to create some really good automations specifically around screen scraping. That’s where it started early on. And so I was at VHA, brought in RPA, and had convinced the CIO to spend the money to get me a license. And from there, we proceeded across the company, and we’re able to do some really great automations. But I was speaking on behalf of automation at a conference and met a woman by the name of Mary Lassity. And Mary Lassity wrote this book. And after hearing me talking and seeing the way that I had approached it, she asked me if I would contribute to one of the chapters on how I was doing it and what I saw coming. And so this is one of my claims to fame, I claim to be a published author, I got to write part of a chapter in this book, and it’s really an amazing experience. I’ll tell you that to tell you this because Marie Meyers, who is the corporate controller at HP, had recently inherited the innovation team. And she was getting our hands around that being a company controller, she was getting her hands around the technology in this thing called RPA. And was really looking at how she can innovate inside of HP. She was at a conference where Mary was speaking and bought the book and read the book on the way home and getting back to Houston. She got with one of the other colleagues that had been at the conference and said, “Man, did you read the book on the way back? This guy Chet Chambers, he’s three and a half hours away. Let’s bring him down and have them look at our department.” So they gave me a call and said “Hey, would you be interested in coming down and helping us out? We’re due on the drain?” I said sure. So I went down there and visited with him and looked at their program. They were using Kryon at the time, and we’re having some real trouble so I worked with them a little bit and came back home. Well about two weeks later they called me again and said “Hey, we were having some trouble. Would you come back?” Sure. I went back down to Houston and one thing led to another and I was asked to come aboard HP. So I came into HP and my full intention was to bring automation anywhere and set it up and run it as I’d been running, but they wanted to run a full RFP. And we did that, we went from eight to five, to three to a head to head and UiPath one the day, they had gotten much further on the complicated tasks, they had found a couple of the traps that I had set and the grit and determination of UiPath and the openness of the tool sold me so we decided to go with UiPath. And from there, we started a team, we were able to develop 224 bots in a little over a year. And it was a huge success, hundreds of FTS worth of automation, hundreds of thousands of hours, and millions of dollars in cost savings across the enterprise.
I’ll tell you, I give you a little bit of history, just so you know who I am, and where I’m coming from. But there’s eight things, eight main lessons that I’ve learned over this last decade. I’m familiar with automation anywhere, Kyron, Pega. And now, at HP, we were using UiPath. And, and these are this is a culmination of what I’ve learned over the last 10 years. So the first line is going to consist of the funnel, it’s a single point of entry. So as you start to release your RPA program into the overall enterprise, you’re going to want to put a push behind it, you’re going to want to draw in ideas from every department. We don’t want to focus just in finance, or just in controllership, we want to work the entire enterprise. But if you have ideas coming in all over the place, then you’re going to tend to lose some of those ideas, and we’re not going to be able to rate and grade all of those ideas in the same way. So set up a single funnel, a single point of entry, to bring all of the ideas in from across the enterprise. That way, it’s not the Wild West. And we review each and every one of those suggestions to see where we’re gonna find the best bang for the buck. You know, in some companies, some parts are automated more than others. But as you look at the enterprise at large, and you let the business users make suggestions on the work that they no longer want to do, on the work that causes them to spend a lot of their time, then we bring those into a single funnel, and we’re able to review them in the same way. So that leads us to esson two. Martha is my mythological person, she’s not a real person, she does not exist. But we’re going to take Martha’s suggestion that came in through that funnel. And say it’s running a report or it’s extracting data from a report, combining it with the data from another report and then posting it out to SharePoint. Well, what data is in that report? And where does that data start? Where does it begin? And you follow that workflow or that data set all the way to the beginning. And from there, you come back down to where Martha’s piece of this isn’t, and you go, “Well, where is the final resting place for this data, or what are the end tasks that need to be done with this data?” So you follow it all the way to the end. And the reason that I say that is that automating 40% to 60% of the piece that Martha suggested is of much less value than getting 40 to 60% automation of the entire process. So if Martha and just her one piece, we can save her a couple of hours if we go from the beginning to the end of this entire data flow 40 to 60% will be multiple FTEs, you know, 40 to 60% of 100 is much more than 40 to 60% of one. And what it does is it lets us take those suggestions in the business, make it real to them early on and let them know that we’re listening, but still be able to grab those bigger values. And that’s by taking the task from end to end. So once we brought these ideas in, and we’ve reviewed them, and we’ve gone into the process, we really need to look at how we determine what is the ROI. And most people when you talk about RPA, they’re very short sighted in the beginning, and it’s not anything bad. It’s just that they haven’t had the opportunity to see what RPA really does. So efficiency is about 20%. So if you take the number of FTEs worth of work that you automate out of the business, times the hourly rate for that geographic region. And the reason I say it that way, is because you know that reducing a headcount in Bangalore, where the cost of the FTE is you know, 10 to $25,000 is much different than then automating one in Southern California where they’re, you know, 170 to 250,000. So it’s the number of hours times whatever the geographic hourly rate is that that’s what efficiency is, and it’s about 20% of the savings. The next piece you need to look at is cost avoidance. If you had to bring in software, bring in contractors, make some sort of change or add some sort of cost to the business to work through this workflow that we’re looking at automating. The next is risk avoidance. Where is it? What do you pay to find where you have made a mistake that has cost the company hard dollars? And if you think that you haven’t paid a fine, go talk to tax, go talk to the Treasury, go talk to invoicing and see where it is that you didn’t claim a two or 3%, early pay credit. Those are things that can be captured inside risk, and what’s the next is compliance. So the Sox and audit guys, and the compliance and the auditors, are going to fight you really hard in the beginning. But what they find out once we show them is that once you automate a process, you can continuously audit that process. And what I mean by that is, for any process, it takes about three weeks in a quarter for the audit team to come in and create an audit over a workflow. Well, in automation, you can quite literally show them the Vizio document that says this is every step that this bot will take in this given scenario. Oh, and by the way, here are the log files that prove that this bot did exactly what it was supposed to for this given record that you want to audit.
So instead of it taking them three weeks, they can literally print it out, print a copy of the map, print the log files for that specific record. And they can do it every day. So instead of it being once a quarter over a three week period, they can do it every day, it’s called a continuous audit. The other thing that the auditors are going to get that they really, really love, is they don’t have to do sampling anymore. Because it’s a bot. And because we’re not hindered by how much data a person can review, you no longer have to check every fifth record, every 25th record, you can check every record. And that is a huge uptick in upside for them. And the fifth is accuracy. Where is it where someone has made a human error that has cost the company money? So those five things together, that gives you your hard ROI. Now, if you extend that out the software savings or the software ROI is employee satisfaction. Now I sit in the boardrooms, and I sit in a lot of events with a lot of C-suite guys to see CEOs and CFOs. Specifically, the number one problem that most of them have, is being able to retain talent. Now if that employee is more satisfied with the job that they do, because the mundane work that they don’t like has been taken away, then they are much more likely to stay with a place where they don’t have to do these mundane tasks that the bots do for them now. The other thing that it does is it gives them the ability to be creative, it gives them the time to experiment to go after the things that they would have always wanted to do for their particular jobs that they’ve never had the time to do. And when you roll these two together, the numbers get so big, so fast. You can’t extend it out more than about three years. Because if you extend the ROI value out five years, people tend not to believe you because the numbers are so big. But I can assure you that every bot that you put in, every bot that has come through that funnel, we have reviewed that task and we have decided to pursue that task from the beginning of the data set to its ultimate resting place. And we apply the ROI values to those bots, every bot that you’ve put in should affect your P&L. If it doesn’t, something is wrong. So the next one is about how we’re getting them in. We’re going to review what the value is and we’re going to determine what that value is. The next thing that you need to do is you need to set up a steering committee. Because you don’t want to be responsible for what you pick. You want to be responsible for how you automate what is selected. So imagine a CFO and their direct staff. So every CFO has a staff of direct leaders. And if you are doing work for the leaders number one, two, and three, if you’re working in controllership, you’re working in tax, and you’re working in sales ops, that’s where all of your focus is, then the people over in Treasury, and direct sales and reporting feel very neglected. And that friction, we call that friction political debt. So you do not want to be in the business of generating political debt, you want the VPs and the Senior VPs to decide what projects you do, based on the value that you’re going to give them with the ROI, you’re going to give them the highest dollar values. But sometimes they’re going to select projects to be done that are of a lesser dollar value, because of the internal political value, because you want to build political capital across these senior executives as opposed to having them against your program, which is political debt. So let those guys determine how to spread this across your enterprise, we’re gonna focus on how we deliver, and that’s one of the keys to setting up a very efficient, very effective RPA program.
So we’ve single funneled, we’re reviewing the process end to end, we’re going to check that out, we’re going to make sure that the steering committee is picking what we do. This is called a PDD. It’s a Process Design Document. And when your solution architects correctly build this document, then the developers spend their time developing using the tool to solve the problem instead of figuring out exactly what the problem is. Now, this sounds like a pretty fundamental thing. But this is a problem in most shops when they set up the documentation there, as his documentation is almost non existent. So if you were to sit with Martha, my mythological character again, and you were to say, Martha, you process invoices, show me all the steps that it takes to process an invoice, she would do that, and she would do that very well, but she would forget about Company X that sends in this funny invoice at the end of the quarter. And oh, yeah, two days before close, we have to process this way. And oh, one other thing, this doesn’t happen often. Those things have to be captured in the as his document and have to be coded for in the to be document. And that’s what this PDD is, this is the document that your developer will use to develop your bot, your developer not should not be going back and forth with the business unit. The Solution Architect should design this in such a way that they spend their time solving problems instead of going back and forth. Now being from Texas, you have to have a story about oil. And I’ve got one. So when crude oil is pumped out of the ground and pumped to a refinery, it’s a chemical reaction that changes crude oil into gasoline. Now it creates a byproduct, it’s a chemical byproduct of this process that made diesel fuel. Diesel fuels an accident, it was a byproduct of creating gasoline. Now doing RPA, I tell you that because doing RPA, there’s a byproduct to doing RPA. And that byproduct of RPA is documentation. So you will have crisp, clean, click level documentation over every process that you automate. You will have how to do it manually, and that can double as your BCP plan or your business continuity plan. Should there be a catastrophic failure with the bots, you can immediately put people in chairs, give them this document, and they can go back and run that process manually, should we have let the people who were doing this process go through automation. More than likely they had been repurposed, it can be called back. But the level of documentation that is created out of RPA is astounding. And when put to good use, it’s just like diesel fuel, you will find the engine to burn it, and it will create value for you inside your company. But this document right here, and we can train you on how to build these correctly, should be sufficient for your developer to code the bots that had been reviewed that had been selected. And that we’ve created the value on. The move to production. So any of you guys that are listening to this, or guys or gals, people that are listening to this, the move to production process is as important today as it was in year one of being an IT professional. So there is no wild west, all code that goes into production is sent through the move to production process, and that’s where the bot’s code is reviewed to make sure that it’s best in class to make sure that it’s not breaking any of the the primary rules, and that we are very tight and very crisp and clean on the way that our production code is is handled. So that move to production is that the information is given or the knowledge transfer is done from the developer to the support person, because developers do not have production access. So as they review the bot, it is approved, the knowledge transfer is done now support owns that bot support is going to be watching 24/7 across your organization through the orchestrator and through the reporting tools that are there. But this move to the production process is critical because not only will developers be building bots, but you will have some high end business users or some sophisticated business users that you will train in RPA development and they will be creating bots on their own. But this move to production is how you make sure that your production bots stay very crisp and very clean, and that support is given all of the information that it needs, so that it can take care of this bot throughout the life of the bot.
Master Level training, if there are two favorite slides of mine, it’s this one RLI. So one of the keys to my success over the years has been master level training my developers. So by that I mean at HP I had UiPath come in and test all of my developers once they were trained, and if they couldn’t pass the test, then you couldn’t be a developer on my team. And if everybody passed the test, then the test wasn’t hard enough. So they ran through testing. We did have a couple that needed to repeat training, but did finally pass so we didn’t lose anyone there, but what we gained is time. And what we gained was speed but Not only speed, it was speed at scale. And that was because my developers were not figuring out how to use the tool. They were master level training, they were certified mechanics, they spent their time applying this tool to solving the problem and not figuring out how to use this tool to solve the problem. Now, that sounds pretty rudimentary, everybody would say train. And everybody should say make sure that these guys know what they’re doing. I’m saying take it a step further than master level training these guys. There should be no difference in the talent that you have running inside your organization than there is with the developers that sit inside UiPath. And this has been one of the key factors to my success over the years: master level train your developers. So let’s go to number eight. Okay, ongoing support. You are in control, you need to think about support before you start. So everybody’s going to get very excited in the RPA programs and your bots as they go live. Everyone gets all excited about the new bot release and saving 2,600 hours this year, $180,000 in savings. Everybody gets excited about that. But that is short lived the way that your RPA program is known. And the way that it will be known across that organization is in the hands of the support people. They are the people that are watching those screens day and night. Those are the people that are jumping in when there’s something that changed or change control missed something and a bot goes down. Frontline is going to assess it immediately. If it’s something they can immediately fix, they will. If they can’t, then they’ll pass it down to level two or level three support. And that business continuity plan that we talked about earlier will kick in. One of the key things that will make a support team very successful is constant updates. If there’s an outage, the support team notifies the business of the outage and supplies those updates every 30 minutes until it’s done. The other thing that is great and really meaningful coming out of support and monitoring is the reporting. So the reporting can be customized. There are C-level people that want to know certain facts about these bots, you can write those reports in Cabana, which is a package that comes with it, or you can write them in Tableau, or whatever the sequel, backend based reporting tool is that you use. But make sure that those reports are written so that you know what your up times are. All the people that have a stake in the bots are given the ability to see those bots run and to watch the performance and to make sure that it’s delivering the data that they need delivered, so that it’s important to them. But there is no more important person in your organization than the person that runs your support and monitoring team. And all too often, support is left as an afterthought. With RPA if you want to be efficient, and if you want to be successful at scale, think about it up front, because there’s nothing worse than putting a bot in production to have it go down and it taking you a week to get it back into production, because you’re having to scramble around and go find the developer that built it, and all of those sorts of things. So if you set your support monitor team up, well, upfront, they will be the face of your RP program to the business because they’ll be the ones that are interacting with them on a regular basis. So those eight things are eight points that if you do well, you can set up an RPA, an internal RPA program inside your enterprise and deliver it scale because fast is good. But fast at scale is better. And I hope that these have been useful to you. If there’s ever anything I can do, don’t hesitate to reach out. And I certainly appreciate you having me here today.
All right, thank you Chet. That’s, that’s great words of wisdom and eight lessons learned on a deep dive on the value of RPA. So hopefully everybody now has some great ideas on what it takes to be successful leveraging RPA. Now we’re going to take a step back, and we’re going to talk about more on the business process side. And we’re going to first start with the challenges of the procure to pay process. So first and foremost, if you look at procure to pay being the green middle slides here, we believe that we need to extend the process, and that too much time is spent in the tactical activity within procure to pay. We’re really now talking about the source to pay through the relationship management of the suppliers and vendors. And that’s one of the things that we’re going to talk about here today. So challenges in that procure-to-pay process: not only ownership of who owns what tasks, but there’s also a high volume of paper based processes. You look at your vendor onboarding process, you look at your invoice process, purchase order process, all of those are very much document based. There’s a lot of non PO based invoices, a lot of manual intervention, a lot of time trying to gather data and analytics on that spend, a lot of organizations may have vendors in the same system, or in the system five or six different times maybe one vendor is in each country or region. And so total vendor spend is a whole exercise unto itself, on what you spend with that vendor. So unpaid invoices lead to communication back and forth, “Hey, did you get my invoice, hey, When am I going to get paid on my invoice?” A lot of these are very tactical activities that you spend your time on today in that process. And really, these are tactical activities that can and should be automated to spend time on more value added activities. The current role of procure-to-pay if you look at these cascading down – majority of the time currently managed and supported by the stakeholders. So 61% of that time is spent on working with technology, 54% spend their time driving financial visibility and providing intelligence to stakeholders basically gathering and reporting data, 50% directly influencing the cash and corporate liquidity trying to reduce the cash to cash cycle, 45% on supplier relationship management, and then 41% on Supply Chain Finance. So if you look at the current state of that procure-to-pay process, and really the source to pay process, we start to now move into more of that extended cycle. The challenges we have today, those activities are manual and siloed. With Sarbanes Oxley and other initiatives, there are reasons why some of those are siloed, but now with automation, we can really start to still keep items separate, but have less siloed information, and certainly less manual activity on the data entry and data manipulation that’s required to do the process today. Right now a lot of the activity is transactional and not necessarily strategic. In polls talking to resources, a lot of their time is spent gathering information for the purchase order processing the purchase order gathering invoices, processing the invoices, gathering the vendor information to onboard, validating the vendor has the correct documentation, maybe minority owned, or woman owned business, or HIPAA compliance or that they’ve completed all their background checks, all of those things are very transactional. So there’s a high amount of data entry and data manipulation for these primary activities. The incentives are typically tied to cost reduction, or some kind of SLA type activity. And they’re focused on very simple outcomes, like reducing your total amount of vendors that you work with. There’s also limited governance.
So in moving to what kind of metrics are measured on, many people on the call are probably measured by the number of vendors doing vendor consolidation projects, spend per vendor trying to negotiate a 2% cost decrease, discounts received, things like that, those first couple are really where we see the majority, what we’re talking about is really now moving more towards that bottom half here, the performance, the perfect order performance, service level and automation percentage really talking about driving business outcomes and driving efficiency of the activities. We’re gonna get into this in more detail. So now moving into that future of the vendor management process, again, elongating that process, and really now moving into more of a value add list of activities. So first, looking at smarter systems that drive greater efficiencies, looking at more deeper agile analytics and reporting, support for full transformation, really collaborating with stakeholders and suppliers’ vendors, and getting rid of those tactical tasks. These are really the goals that are essential for having that more future driven process flow. So as we talked about the current state now, moving towards the way forward. So moving from manual and siloed, to more automated, and you heard chat talk in lesson two about the full process flow, so not just looking at a task, but the full automated tasks. So that procure-to-pay source to pay kind of the full lifecycle of vendor management, paving that road. From an automation standpoint, to take away all of the manual activities, and allow you to really take your items and tie them to corporate objectives and the bullet point here. We also become more analytical and data driven. So as we have automated processes that are tracking all of the data, you have this wealth of information that you’ve never had before. And you can start to have more predictive analytics, more artificial intelligence machine learning, added that now will start to give you better insights into data that you typically haven’t had the time or the ability to forecast and to look at previously. Incentives tied to business outcomes: so really starting to partner with your vendors, or suppliers on mutually agreeable outcomes. There’s the talk about the gorilla and the elephant, where the 800 pound gorilla is the organization and you dictate terms to your suppliers, and you kind of pound on the table and demand certain terms. Well a vendor is the elephant and they never forget. And so they come back. And they look at all those terms. And they start to in their contracts, put in risk avoidance. And so now both sides are fighting over, in essence, table scraps, trying to protect their own and get the percent here, percent to their side. In this future environment, we’re really driving much more towards mutual incentives that tie to true business outcomes. And really having a more holistic view of vendors across the regions, across the business units, across your services, so you’ve got that one true visibility of vendor spend. And also, of course, enterprise governance. So really trying to move this from tactical tasks on your left the current state siloed tactical manual tasks to holistic, mutually agreeable, future state activities that are much more meaningful and driving your overall business objectives. So you start to become an advisor to the business and provide meaningful difference to the overall goals and objectives of the organization.
Now, talking about how we can tie these two things. So Chet talked about RPA, and he gave you all the value of how to be successful in RPA, we just went through the problem state and the future state of the vendor management process, the holistic vendor management process. Now we’re going to talk about how RPA can really get involved in these activities. So first and foremost, taking a step back, what is RPA? So here’s the definition, from the Institute for Robotic Process Automation and Artificial Intelligence of which we are a member. You can kind of read that definition on your own, but we really break this down into a couple different tasks. So first, RPA is attended by RPA. This is where I think of a call center or help desk kind of model, where you kick off the robotic process automation, you initiate that, and it controls your desktop and runs some processes, as if you were doing it just, you know, better, faster, cheaper, than you would do it. There’s an unintended RPA. And this is much more of a batch process that runs in the background or the lights out process. That’s kind of a high volume does it on its own similar to you running batch activities in your PO or AP system today. There’s also a hybrid process. So you can look at things that run unattended and then come back to you and require you to enter information. So think of going out and sending some files to the bank. That may be fully automated, but your bank may require a key fob authentication. So it would initiate that process up to logging into the bank. It would then ask you to do your process, and then it would continue the process along. And then there’s also cognitive RPA. And certainly organizations like UiPath are well known for being a leader in this space. Cognitive RPA starts to add more of that intelligence into the bots where it’s not just doing the manual tasks to actually starting to think and process based on doing this over and over and over again. So it starts to look at unstructured data versus only structured data, and it starts to improve as it continues to run these activities. That’s a little bit of a background on RPA, where you typically and similar to Chet’s activity, or lesson learned about what kind of where RPA is valuable. These are the five headings that we talked about. And I’ll just very briefly go through this and some of its repetition, but we look for hours back to the business, or productivity gains. So when you’re thinking about ideas on where RPA is gonna fit in your source to pay procure-to-pay process, you think about what are the manual activities that you do over and over and over, if you could kind of push a button and they’d be done, you would have X amount of hours back to the business. And also the fact that this can run 24/7. So if you have shift workers, where you’re doing procurement in three different shifts, 24 hours a day. Now, this is something that now you can do with robots 24 hours a day. Also cost savings: it is certainly cheaper to run a process than to have three shift workers. So you can empower those people to do more meaningful work versus do the tedious work. Risk Management: Chet also talked about that, where is it extremely important that you have 100% compliance, and 100% accuracy. Innovation really talks about empowering those people to now go out and do more meaningful activity, and work across functions. And then employee engagement: we did a survey across multiple organizations, a couple of 100 organizations, I don’t remember the exact number off top my head, and 95% of those employees felt like they made a meaningful difference to the organization. But those same 95% that thought they made a meaningful difference to the organization felt like the day to day activities they do do not make a meaningful difference. So think about that the average person feels like they make a difference. Yet, the tasks and activities that they do every day do not make a meaningful difference. And almost all of those employees felt like they would be more engaged if their activities were more meaningfully empowering to the organization.
So some common examples across that procure-to-pay process. Again, the types of activities you’re looking for are manual, repetitive data entry type activities. Ideally, they touch multiple platforms. And the reason here, they don’t need to. But if you use SAP there and you have data in Excel that needs to be loaded into SAP, well, there are probably many ways that you can load that SAP or that Excel data into SAP. It doesn’t have to be RPA, it can be RPA, but certainly the more applications you touch, and the more things you need this process to do, the more value RPA is going to provide. It certainly needs to be high value, because it’s got to justify the cost of creating this activity, this RPA. So the more high volume, the better. Something around that quality control, data validation, and certainly it needs to be business rule driven. So it can have conditional logic, that kind of if then, else, type logic. But if you can script it out, if you can write it out, what it needs to do, then you can automate it. So some examples of use cases here. Talk about vendor selection and onboarding, we’ve we’ve mentioned that AP invoice scanning, so receiving an invoice in an email, scanning that invoice and then entering that pertinent information from the invoice into your AP system, expense receipt, an entry, taking a picture of an expense receipt, submitting it and having that go into your expenses system and entering that for future payable. Shipment status: reaching out to three PL applications logging in and gathering information about where your product is and communicating that status to maybe your end client. Order creation sync, synchronizing that what we call swivel chair data entry, where you may need to enter data into an order management system. You may also enter that same order into a CRM system or into other applications, warehouse management systems. Having that data entered once and then repetitively entered into those other applications isn’t easy automation. Obviously anything that’s in an Excel or a CSV file and importing that into a system is no better for automation. Competitive price comparison: you can go out and look at your vendors and look at competitors of those vendors scour their website and material numbers for competitive price information, making sure that you’re using the right vendors in the areas that make the most sense, providing that 24/7 support, again, allowing the bots to kind of run 24/7, any kind of helped score information around a payment info automating those scripts where you’re sending your vendor information that “Yep, I’ve received the invoice I’ve entered the invoice he’ll be paid on this day,” sending them another one, two or three days prior to payment saying you’re going to get paid on this day or I just cut the check, any of those kinds of communication is going to save you having people call you and ask for that information. Vendor analytics and scorecards: another great example of gathering data to create scorecards for your vendor, and we’re going to talk about that more in a minute. And then supply and demand planning as well. All great examples, and use cases here.
So real quick, where are others on the journey? This is from APQC, one of the market leaders in best in class information. They provided some information on RPA. So if you look at the grid on the left, you’ll see that just under 20% of people have not even started their RPA journey. So if you’re on the call, and you’re feeling like wow, this is all new information. To me, you’re not alone. But certainly the majority of the people are kind of in the early planning slash piloting stages. That’s where more than 50% of the people are. And then you’ve got another 20+% at the end here, where they’re already turning this into more of a programmatic solution. So really, kind of across the board, you’ve got people across all areas. The challenges that you see: similar to the lessons learned, 48% of the people lack process foundation. So it’s very, very important that you do your process discovery and you don’t automate a bad process. So you really want to make sure that you’ve got a strong foundation on how the process should be run. That naturally goes into the next one streamlining or transforming the process for automation, doing process improvement prior to automation, making sure that you’ve got a center of excellence that’s really driving a guiding strategy for automation, and making sure that you’ve got strong executive sponsorship on how these decisions and business rules and models should occur.
So how do we make everything that you’ve heard? How do we make this sustainable? How do we create a very successful solution going forward? We have what we call advanced process intelligence. And so at Ashling Partners, this is the process that we recommend to all of our clients, and all of you on the call. So first and foremost process discovery, really ensuring that you understand the process, you document the process, very similar to what Chet was talking about, you map that out, you look at all the exception handling that needs to occur, and you understand how that process should work. And you make it as efficient as possible prior to automation. Then you add automation. And automation can be multiple things, it’s not just, one hammer looking for a nail. Maybe it’s better to create a report for so and so or maybe it’s better to build an interface. But automation is really going to be key where we call out those steps on the prior slides. And then most importantly, you want to build an automation Center of Excellence and have analytics around your bots, the utilization of the bots, and create a model for sustainability, to really enforce a continuous process improvement. And all of these should be focused on what we call the new BPO business process outcomes. So all of this really needs to be centered on what are the objectives that we are trying to accomplish by improving this process, both from an efficiency standpoint and automation standpoint, that’s really first and foremost.
And so if you look at how we define the process flow on how to do this. First and foremost, you need to define your corporate objectives and business outcomes. What are we trying to accomplish? Are we trying to reduce costs in this process, therefore be more efficient and give hours back to the business? Are we trying to get better visits ability on our vendor spend? Are we trying to analyze our vendors into what we consider core and differentiated? So core are items that keep the lights on, they’re not strategic to the overall organization, but they are necessary. And then differentiated is more of that innovative area that furthers your differentiation over your competitors. So, for your typically consolidating vendors, you’re typically looking for cost out. Differentiated, you’re typically bringing in boutique vendors, you’re trying to drive innovation, you’re trying to educate your business, your internal customers, if you will, on the types of things that vendors can do and educate them on innovation and ideas. That’s all corporate objectives. So what are we trying to accomplish here? And then we’re going to align our KPIs or metrics to those corporate objectives. So how should we measure ourselves, not just on vendor spend, but what are other ways to improve how we measure ourselves and our value back to the business. Then we’re gonna do that process discovery of the current state, and map out that process, we’re going to establish and validate our sourcing strategy to align with this process and with our KPIs. We’re going to develop that future state that includes where automation and where efficiency can be gained. We’re going to benchmark those metrics to best in class, and that’s where we use APQC as those best in class, benchmark metrics. And then you’re going to quantify the high level value, in essence, build the business case of saying, okay, here’s where we are compared to our metrics are the benchmark data, here’s where we think we can be with the process improvement and automation, here’s the value that that’s going to drive back to the business that’s leading to these measures. And here’s obviously the cost to do that. So you’re building that business case, based on all these business cases, you’re prioritizing the automations that you should be doing. You’re going to then build those automations, and then you continue on, continue on. So that’s where we really believe that your KPIs need to be continual KPIs, not necessarily things like headcount reduction, because that’s a one time KPI. So if you look at, again, APQC is our partner organization, we use them for all of our benchmark data. For example, information that they would provide, and more of that continuous process improvement. You look at financial indicators are typically lagging indicators, they’re kind of after the fact indicators. You’ve got qualitative indicators, you’ve got efficiency indicators, you need to come up with the ones that make the most sense for you. But here’s some examples that APQC would provide and an example kind of benchmark data. So from a cost effectiveness kind of a cost model, the total cost to procure materials and services per 1000 in revenue, which just allows you to standardize based on your size. Compared to other organizations, here’s kind of your top performer, your top 25%, your median, and then your bottom 25%. Same thing with the number of headcount it takes to perform that activity, which is more of a staff productivity standard. Process efficiency: how many orders are approved electronically? And then cycle time: how long does it take to process? So I’m not going to read all the slides here, but these are examples of best in class metrics that you would use to go forward. And to look at things more in a continual process improvement. You could say, if you’re a bottom performer, you’re going to get to median. And then for meeting, you’re going to get the top performer, how are we going to do that? We’re going to improve our process model, the process mapping, we’re going to come out and we’re going to look for efficiencies, then we’re going to look for automations. We can try to expedite that process significantly and reduce that total cost.
Here’s an example of what we call the Ashling Automation Analyzer. This is where we go into detail on an example procure-to-pay process, we map out that process, think like a Vizio process, if you will. And then we store metadata at each one of those boxes within the Vizio process. And then we can analyze it to say how many participants touch the process and how many times do the participants touch the process? So in this process, flow procurement touches it five different times. The business touches it four different times, the AP team touches it three different times. Is that necessary, do we need to provide that data so many different times to the same groups? Are there different ways of improving this process? Or can we automate a couple of those steps?
This is what’s called the Heatmap. So these are all of the activities in that full vendor management source to pay process that we talked about earlier. And again, you need to do this for your organization to make it specific to you. But just to use this as an example. The blue boxes, the ones outlined in blue are ones that would have a rather low impact to try to automate. So for example, negotiating contracts, you’re not going to automate negotiating contracts, that’s going to be something that’s left for you to do. But that’s more of a value added activity. If you look kind of in the middle here, there’s a lot of red. And that’s typically where people spend the majority of their time creating requisitions, approving requisitions, generating the PO, releasing the PO to the vendor, recording the bill of lading when the when the inventory items come in, receiving an invoice, paying an invoice, matching that invoice to the purchase order, approving payments, all of those things can be automated, preparing vendor information validating terms are met. If you go all the way over to vendor management, those are things that can be automated. So these are examples of high impact where we can take those manual activities away from you. And then you can spend time on doing the vendor performance analysis, monitoring and managing supplier information. Working with your business, on purchasing requirements, purchasing strategy, inventory strategy, the supply and demand capabilities. All of those things are where you can add a lot of value to the organization. And you can and should be spending time and allowing automation to take away a lot of these tactical activities. Again, this is an example one, you should create yours, and work with the business on where automation best fits for you.
So with that, we’re going to wrap it up here and thank you for your time. You can see our contact information here if you have any other questions or comments. We look forward to following up with all of you individually, and hope that this was valuable for you. If you have any other follow up questions, don’t hesitate to reach out and we will gladly talk about this because it’s something that we’re very passionate about and I look forward to working with you. Thanks for your time.
Length of Webinar: 52:39 Minutes
Originally Presented: November 23, 2018
At Ashling Partners, we work with leading technologies in the RPA and broader intelligent automation markets.The result is a more efficient, agile, and engaged workforce and a business-outcome driven company.
Ashling provides education on the evolving automation market opportunity, implements and operationalizes your process and technology capabilities, and structures a sustainable business model using four major pieces of intelligent automation.