128K-Member-Banner.jpg

A Parallel Universe

So What's the Deal with RPA?

Contributor: Barbara Hodge
Posted: 01/09/2017

RPA – to be or not to be?

Quoting Hamlet (Act 3, Scene 1, if you are wondering) will probably only win me points with the literati amongst you. But we are just about to announce our annual State-of-the-Shared-Services-and-Outsourcing-industry survey report, and I've been intrigued with a few findings (you'll be able to read all about it yourself, later this month, when it’s posted online). One thing you'll no doubt identify with is the fact that, to date, and despite all the headlines we've seen over the past two years, process automation, or to use its "sexier" descriptor, robotic process automation, is still ranking low on every front.

This is intriguing to me. Not only has the market not implemented RPA (60%+ of HR, nearly 50% of Finance and 50% of Multifunctional respondents say that they have done nothing yet) but the vast majority confesses to not understanding its value.

From where my team and I are sitting, this is strange to comprehend and stranger still to believe. After all, at its simplest, RPA offers a human-touch-less workflow across user interfaces that does away with routine, mundane – and let's not forget: more error-prone, handovers. And perhaps that's the problem: Automation is just too much of a common sense idea.

I think a key issue might be that many organizations may be benefiting from RPA without necessarily knowing it. There are literally dozens and dozens of small specialist process automation providers whose business consists purely of feeding RPA into BPO provider solutions. So if you're getting a "better" result from a long-term provider, this may be because they've implemented some pretty smart fixes on the backend. Or, if you haven't put adequate performance improvement targets or gain-sharing clauses into the contract, you may have just become a higher margin customer, without the advantage being fed through to you.

I've spoken to plenty of these smaller-scale developers over the years who are frustrated at the lack of recognition by end customers. "They don't recognize that the amazing new solution they just bought into from a big-name provider is built entirely on our capabilities. We just can't make those connections directly," one explained to me last year.

The other issue may be that RPA providers, who obviously have a lot of educating to do, are targeting the wrong folks in house. According to our survey results, the people most likely to buy into and understand the strategic value of this kind of automation are the functional heads – HR and Finance leads. Targeting the CIO is fairly ineffectual by comparison.

Oddly enough, it's not budget availability that's preventing more RPA from being implemented, at least not according to our survey; it's not even the lack of a watertight business case (although IT tends to cite this more – of course, they would). The real issue seems to be lack of resources to allocate to the project, and competing investment priorities.

So these RPA assessments aren't even really getting off the ground.

Earlier today I chatted with Nick Andrew a founding partner of Virtual Operations, which specializes in what he prefers to call "intelligent processing" (by this he means deploying intelligently, i.e. holistic, strategic, and with all the tools available, as opposed to simple 'artificial intelligence'). Nick's background includes Alsbridge, EquaTerra, and Everest Group, so I’m going to trust his judgment when it comes to sourcing.

The problem, he says, is that companies are touting enormous reductions in FTEs without much (any?) due diligence. In addition, RPA alone can't really achieve the kind of results that are being touted – and this could lead to a dangerous backlash as disappointed consumers make themselves heard. Effectively, RPA is a piece of the solution and potentially can drive an enormous transformation in our time. But it depends on taking a fairly end-to-end, or holistic, view of the enterprise. Yes, you get results processed by process, tool by tool, but these are generally going to be chump change compared to what could be done with a wider mandate. What's key is that to get the real bang out of RPA requires far more than a single tool. You need artificial intelligence, OCR, and a whole host more, that can decipher and integrate unstructured data as well as structured. Success depends on embracing digitalization to a far greater extent than many of today's functional head are comfortable doing.

And therein may lie the real problem. Conventional CHROs or CFOs will see the benefits to their function only, and measure the impact in terms of reduced FTE requirements or costs. The real returns, Nick believes, come from leveraging a technology-driven Center of Expertise that incorporates intelligent automation and pushes this out across the enterprise. This won’t be the same as old-fashioned IT technology, he explains, but is based on recognizing that new technologies are replacing humans. The best person to lead such initiatives may well be a new-age CFO (who’s focused less on transactions and more on ‘value’) or a Chief Operations Officer. Certainly, it requires someone with a lot of clout and a long-term view, as you might be looking at a 5 to 8-year time-frame to reap the full effect of change (they will also need to get the CIO aligned, as IT should always be involved in managing governance).

For those who are interested, it may make sense to start with a tactical approach, which can fund the straegic approach: i.e., start with RPA for low hanging fruit, while working on some of the more transformational pieces in the background. By the time you have your business case together you'll be able to point to the success of your pilot program.

For now, we will have to wait for more large-scale case studies to appear in the market to lend credence to the argument in favor of process automation. From the conversations I'm having, I think we can expect to see far more of these within the next two years.

Contributor: Barbara Hodge