Among the many technologies with potential to completely alter the workplace, clearly robotic process automation, RPA, is a leader.
RPA enables companies to save precious staff time by automating many business work processes. It can interface with software – mimicking a human – and perform reams of manual work. Those hours and hours of repetitive tasks that employees do can be performed faster, and more cost effectively, by an RPA robot. Think of RPA as “pragmatic artificial intelligence.”
The RPA bot are programmable at many different levels and are gaining sophistication with time.
Yet RPA is still on the cutting edge. The market is growing rapidly – revenues are forecast to grow at a scorching rate – and RPA tools are still seen as “new” by many enterprises.
To provide advice on best practices to maximize an RPA deployment, I spoke with an industry expert, Eric Tyree, Head of Research and AI, Blue Prism.
Best Practices for RPA Deployments
1) Where are we with the RPA market now? It seems that there’s great interest in RPA, but it’s not yet fully mainstream. True?
2) When companies start to deploy RPA, what are common challenges and set backs they encounter?
3) What are some best practices for companies to ensure maximum success with their RPA deployment?
4) Where is RPA technology headed over the next few years. How can companies that want to deploy RPA – or are deploying it now – prepare for these changes? Â
Listen to the podcast:
Watch the video:
Edited highlights from the interview – all quotes from Eric Tyree:
Is RPA Fully Mainstream or More Cutting Edge?
It’s definitely past the early adopter phase. I wouldn’t describe as mainstream yet.
I think it’s reached the point where most companies are aware of it to some degree. I was first seeing it five, six, seven years ago, actually trying to use it in my previous company, so about a year ago. So I think it’s well known, in other words people are aware of it, they know it’s a technology that’s out there, but I don’t think it’s seen as a mainstream standard tool kit as part of the overall digitization piece. I think it’s beyond early adopter, but it’s definitely still emerging, is the way I would describe it.
Challenges in RPA Deployments
I think the number one problem is it’s too low down on the company, so it’s used by a small group of people in a small business unit, they don’t have the support they need, it’s not sanctioned by senior management, they don’t have the budget and supports they need.
What’s true of automation is what’s true of digitization. The key to success is that you’re doing two things. One, it’s goal-directed, it’s very clear what the business case is for, or what you trying to achieve is measurable. But also that you’re looking at whatever you’re trying to automate, just like with digitization, you can’t just take what you’ve currently got and automate it, ’cause all you’re doing is taking your existing spaghetti ball and automating the spaghetti again, which isn’t really giving you much value.
Whereas successful [deployments] are bit more thoughtful, they really step back and think, “Okay, if I’m going to take my mortgage desk and I’m going to apply automation to it, what I really should be doing, is not looking at my processes, copying them, and just taking the manual working and replacing it with the digital worker.” What you should be doing is saying, “Okay, if I’m gonna redesign mortgage processing from an automation first perspective or a digitization first perspective… ” [Instead] you start with a blank slate and you redesign the entire thing,
And you think, “Okay, where is automation’s appropriate place? Where does human work have its appropriate place.” And affectively what you’re doing is you’re rearranging the way in which work gets done that’s getting the best out of robotics and the best out of people.
It’s probably part of a bigger company strategy, and that’s when you get to success in [RPA].
Best Practices for RPA Deployments
So I think it depends what you want to achieve. So I’m assuming that you’re looking for something transformational. Obviously, if you wanna go ahead and do something small, it’s a different set of priorities.
But let’s just say this is something big and you want to start with your mortgage desk and you’re gonna move to other back office systems, and you’re doing a big automation plan.
The first thing you need to do is make sure you understand what it means strategically. So why are you even doing this? As a bank, are you doing this because the market demands five-minute mortgages and you have to start providing it, and this is all about customer service, modernizing the shop front? Is this something else?
Your problem now is, how do I have to redirect my human capital from non-revenue generating to more revenue generating work? So again, you’re thinking top-down strategy, so what am I gonna do with automation?
In that case is I’m gonna attack non-revenue generating work so I can free up those people to spend more time working with clients and doing the value add. So again, it starts with that strategic alignment. Thats’s the most important thing, without that, it’s gonna sit there and fester as a small project and never get anywhere.
The next piece is to set up your center of excellence. It does take some skill and some knowledge. It’s the kind of thing that a power user can build, you don’t need a PhD in robotics to do it, you need to be the person who is comfortable with macros [RPA is low code].
All I argue is that the bigger your ambitions are, the more senior and the more aligned you need to be with strategy. I think that’s true of any project, particularly if you’re gonna be playing around with the company’s operations.
You want to know that you’re aligned and your boss and your boss’s boss, and your boss’s boss’s boss all know what you’re doing and agree to it.
The Future: Humans and Robots Melding
[With RPA], you’ll respond, and you won’t realize that the person who messaged you was a robot.
And that’s already happening, but we wanna take that to the enterprise scale, which is when you redesign your mortgage desk that’s what you’re designing in, you’re literally saying, okay, if it’s all the sort of generic mortgage applications out there that are a low risk, they get swept through by robots.
Robots are then looking “Ohh this one looks a bit borderline.” It’s going go and ask Joe over there what he thinks, he might ask Susan as well what she thinks, Susan responds. She doesn’t realize that the person who asked Joe in the first place was a robot, it doesn’t matter. It’s just an IM that’s come through Teams or Slack, it’s just a message.
And that’s the trick, and I think the real future of it, it’s not just this enterprise scale. Robotics is increasingly passing the Turing Test, people don’t realize they’re talking to them.
Chat bots are a great example. The trick with chat bots is how do you triage between humans and a robot without someone knowing? So a really good chat bot flips you to a human and back and you don’t realize that it’s flipped.
So it’s that triage process is the key, if you get that very quick, then you get a great customer experience because a robot is handling all the routine stuff. Humans will be able to dive in solve problems that the robot can’t, and the consumer is just getting a great experience and they don’t realize that it was two different types of entity that they were interacting with.
