In the last 12 months, private investors flooded the robotic process automation (RPA) sector with cash, even as equity markets were marking incumbents down.

That new money might be evidence enough of the emergence of a robust, incontestable business idea: that code is faster, more accurate, and much cheaper than the humans it replaces. But, of course, the truth is more nuanced.

Even advocates like KPMG — in a paper for Chartered Accountants Australia and New Zealand called Embracing robotic automation during the evolution of finance — see some initial speed bumps. “The growth of ‘digital labour’ will affect organisations for many years to come. In the short term, some organisations may struggle with disparate and uncoordinated automation initiatives, as well as fragmented underlying IT systems and applications. There will be continuing uncertainty over where to best start, when to, and how to invest in automation.”

And KPMG argues that talent management challenges will need careful consideration.

But ultimately KPMG argues, “Within the automation choices available to CFOs that have recently emerged, RPA has garnered significant market attention, and for good reason. Our research suggests that the benefits of adopting RPA in finance go way beyond cost reduction bringing improved control, faster processing speed, better data, quality and happier finance team members freed up from mundane tasks for interesting and value-add work.”

Yet, the very fact that industry leaders in the space are trying to step quietly away from the RPA brand — to redefine the sector as “Intelligent Automation” — hints at some potentially deep-seated issues.

Violent agreement

While even its critics acknowledge that RPA has benefits as a tactical solution in the right context, they argue it has been massively oversold. Over the last two months, Which-50 has spoken in depth to both critics and proponents, and the most remarkable outcome is how much they agree with each other on some key points.

Those of a skeptical bent say the technology has largely been applied to low-cost jobs, it doesn’t scale well beyond the enterprise, and the cost of maintaining all those robots is often higher than customers initially anticipated — though that’s often as much an issue with RPA implementation partners as with the technology providers themselves.

Where critics and advocates diverge is on a point of perspective. Critics argue the situation is irredeemable, and that companies such as banks and financial institutions would be better investing their time and capital on truly transformative initiatives.

Advocates amongst the customer community, however, say it’s simply a matter of market immaturity. They argue that any problems reflect the normal compromises of early adoption (though it’s not technically new technology), and they expect the expertise they develop now will help them steal a lead on competitors later.

Customers are also generally less squeamish than vendors about the headcount reductions which are often used to justify projects.

At the Blue Prism conference in London earlier this year — which we attended as guests of the company — executives like CEO Alistair Bathgate were at pains to argue that RPA is about much more than headcount reduction.

Yet almost every customer presentation we witnessed framed the business justification almost exclusively in terms of staff cuts.

The money trail

Three companies dominate the RPA sector: London based Blue Prism, a listed company, and two US-based private ventures, Automation Anywhere, and UiPath. Founded in 2001, 2003 and 2005 respectively, the companies have all taken their time to become overnight successes.

But now the sun is shining, and all have been busy refilling their bank balances with investment capital over the last year.

Most recently, UiPath revealed it has secured $US568 million in series-D funding, which values the business at almost $US7 billion.

For its part, Automation Anywhere has raised a similar amount — $US550 million in series-A funding — while Blue Prism, the most mature of the three businesses, added $US120 million last year.

As a public company, Blue Prism also has to contend with extra scrutiny. That much was clear during its conference, when it had to address questions about the extent of its R&D spend.

Investors suggested the company’s R&D budget looks lower as a percentage of revenue than many comparable technology businesses. The company disputes this characterisation, suggesting it is a matter of definition.

The company also saw its share price go into steep decline between September 2018 and January this year, when it fell from over $US33 to just over $US13 — although it has staged a strong recovery since then and now trades at just over $US24.

Not impressed

Share market investors aren’t the only ones who have had questions about the wider category in recent times.

The skeptics we have spoken to about RPA list a range of concerns:

  • Many BPM vendors argue that RPA negates the need to upgrade core banking/finance systems (or delays the need for an upgrade). The problem, however, is that many of those core systems do not support contemporary computing models and services like APIs, and this delay can be problematic at a strategic level.
  • Many of the tasks being automated are low-level functions being carried out by people in low-cost labour markets meaning the payoff is marginal.
  • RPA vendors deliberately obfuscate the reality that the code and the automation need to be maintained over the years, and the cost of this maintenance easily soaks up the savings. That’s because the people doing this work tend to be highly paid consultants.
  • RPA is pitched as a path to AI and machine learning when it’s not.
  • The financial performance of RPA vendors in recent years suggests the market is turning against the model.

Perhaps the people best-placed to comment are not the vendors, but their customers. At Blue Prism World, we spoke to Tomáš Přibyla, Director, Operations Specialized Services at CSOB, one of the largest banks in the Czech Republic, and a customer who delivered a case study on the main stage of the conference.

Přibyla also has a wider perspective of the RPA market, since CSOB also uses solutions from other vendors like UiPath in some of its other operations. He describes the various solutions as commodities, and he sits firmly in the camp that says the problems that exist today will be overcome as the market matures.

We started by asking him to address a common theme of many customer presentations at the event: the difficulty they faced scaling their automation projects across the organisation.

“It’s quite straightforward,” he said, “It’s about the openness [of managers] in the organisation to change and for sharing cases that might endanger their existence. Those were the typical hurdles we were tackling in the banking part of the organisation but also in other parts of the company.”

If you are a manager, he said, you need to be open to sharing and showing your processes, their potential for savings, for optimisations and for integrations.

Many managers can find this challenging, he suggested. “Culture changes don’t just relate to the process managers but also to their managers.”

The culture change piece is core, he said, and technology is not the impediment. “There is no technical limitation with Blue Prism, or with UiPath, in the other parts of the implementation.”

Blue Prism was the best fit for the CSOB pilot, and by the time they were looking at automation in other parts of the company UiPath provided more favourable commercial terms which made it more appropriate in that instance. Specifically the fact that it offered open community support, he told Which-50.

On the cost of maintaining the bots — another common criticism of RPA — Přibyla acknowledged that this is an emerging market and there is a price penalty for first-mover advantage.

“The cost of the licences now is going to be different in three to five years. But if you want to jump on the wave and be there, this is the cost you are paying. But unless you adopt it now you might come too late to it .”

He also said labour costs remained quite high in the markets CSOB operates in, and the competitive advantage of shifting to automated processes stacked up favourably.

But he said the FTE reductions at the bank typically did not translate directly to job cuts, as CSOB is currently growing and many of the roles were able to be redeployed.

Advocates abound

It would also be wrong the suggest that all customers are critical of the technology. Many say they can point to strong evidence of success.

One such customer is Stewart Munro, a senior executive with the Australian Department of Finance — a Blue Prism customer — who outlined the benefits the organisation has accrued with its automation projects, and how it overcame skepticism and some fear to do so.

He addressed directly the issue of the impact on employment. “When we started this journey, there was a lot of talk around people and the fear their jobs were being lost.”

Unlike many of the finance and banking institutions that spoke before him, the Department of Finance was able to address this by stressing that no one would lose their job in the next year through automation.

“We’re very confident we can do that. So what we said is you will want to engage.”

One of the ways Munro’s team were able to engage staff was to sell the automation as digital assistants, rather than robot replacements.

“We started with some clear articulations of what does this mean for them as individuals. We talked about what we were going to do from a management perspective.”

This approach led to a strong improvement in employee engagement, he said.

Previous post

Paradox of Choice in Digital: Marketing, Technology, Data and Customer Experience

Next post

Accountability will be key to adtech success, says CtrlShift CEO

Join the digital transformation discussion and sign up for the Which-50 Irregular Insights newsletter.