Design thinking has helped my team reframe problem-solving to innovate for customers - but it hasn’t always been smooth sailing. Consider these five lessons learned.
Robotic Process Automation (RPA): 5 lessons to learn early
Consider these lessons – ideally, before you get started with Robotic Process Automation, experts say. This prep work will help you – and your bots – reach RPA success
4. RPA may be a breeding ground for shadow IT
Love it or shove it, the term “shadow IT” does have at least one useful feature: Just about everyone in IT understands that the phrase refers broadly to other departments acquiring and running various technologies on their own, typically without any approval or oversight. RPA could power a new wave of that practice. That’s because, again, RPA vendors have a bottom-line interest in making their development and management tools as easy to use as possible. With some, no coding or other real technical skills are necessary, the idea being that a business analyst or other non-technical people can build a bot.
On the one hand, this is a good thing, because you’ll need folks from outside of IT involved in your RPA strategy. Otherwise, you might not have the proper visibility into how a process currently runs.
“Trying to automate processes you don’t understand is a path to failure,” says Antony Edwards, COO at Eggplant. It’s a common RPA mistake; avoiding it typically requires cross-functional expertise.
Chris Huff, chief strategy officer at Kofax, notes that no one understands a business process – including what’s wrong with it – as well as the business users responsible for it as part of their day-to-day job. So, yeah, those folks should absolutely be a key part of your RPA team. But that doesn’t mean they should necessarily operate without any IT collaboration or oversight. This is an area where the ease of use that multiple RPA platforms tout can backfire.
“This can also lead to the business buying, implementing, and attempting to scale RPA without the CIO,” Huff says. “Very quickly the business runs into the IT wall, which tends to occur when transitioning into full production.”
That’s the point, Huff explains, when a department begins hitting up IT for user IDs and other credentials for their bots, or making system-change requests so that their RPA bots can access different IT stacks across networks. This is also the point when a business user realizes: Oh, there’s way more to this than I realized. In other words, ease of use, even with the best of intentions, can produce a pain in the you-know-what.
“The bottom line is that when the business [user] gets over their skis and fails to include IT in the early planning stage, bad things happen,” Huff says. “The program typically stalls out and goes into hibernation until an executive bridges the IT and business divide and moves the program along in a co-share and ownership manner.”
There’s a balance to be struck here. Zero IT involvement, as Huff argues, is almost certainly going to cause problems. Too much IT involvement could produce a maintenance boondoggle of the sort that most strategic IT leaders are actively trying to get off their plates.
“Ensure the business or functional experts can update the RPA without going back to IT,” Edwards advises. “Otherwise, the lag will be too slow and RPA will always be useless.”
5. Sustainable success requires structure: Consider a COE
Perhaps you’ve picked up on a theme here: Putting a bot into service might be relatively straightforward, but achieving measurable, sustained results with RPA is not.
Huff from Kofax sees plenty of organizations learn this the hard way, usually when they start to ask why they’re not hitting their goals or why they are otherwise failing to gain traction. The lesson learned: You’ll need some kind of sustainment model, Huff says, if you want to achieve the kind of benefits that are generating so much heat around RPA in the first place. Huff recommends the Center of Excellence (COE) approach for your RPA program. Another term for it, one that gives a nod to the importance of treating bots as a new class of worker, is “digital management office.”
Adobe CIO Cynthia Stoddard recently shared with The Enterprisers Project how her team scaled its RPA usage rapidly – and got crucial buy-in from other departments – with a COE. You can learn from her experience.
Whether you call it a COE or another name, suffice it to say you need appropriate structures in place – the sooner the better – to drive long-term RPA success.
“The key here is to start thinking about your COE during the initial pilot and proof-of-concept,” Huff says, “because it takes time to get the right executives bought into the business model changes required when developing a new RPA-enabled digital workforce.”
[ How can automation free up more staff time for innovation? Get the free eBook: Managing IT with Automation. ]