Prof Leslie Wilcocks of the London School of Economics says someone once introduced him as “the pope of robotic process automation.”
“Which, of course, I had to deny, as a self-effacing Englishman,” says Wilcocks, one of the top researchers in the field of knowledge work automation.
The moniker is a reference to his extensive research on RPA, which led to his co-authoring with Mary C. Lacity three books on the topic, Service Automation: Robots and The Future of Work (2016), Robotic Process Automation and Risk Mitigation: A Definitive Guide (2017), and Robotic Process and Cognitive Automation - The Next Phase (2018).
The fourth book, Becoming Strategic with Robotic Process Automation, is out later this year.
As he tells, however, a breakfast forum in Wellington last week, when it comes to RPA, “The only thing I am likely to be is St Christopher, the patron saint of travel.”
“I am here to [help you] navigate a little more safely than you would otherwise do.”
Wilcocks says he personally looked in detail 420 deployments of RPA to get insights on what works and what doesn’t.
“When I first looked at RPA, I thought this was a no brainer,” he admits.
“RPA takes the robot out of the human,” was a phrase that was widely quoted from his first book.
He also shares “not a nice thought” to the early morning audience. “Children were the robots of the first industrial revolution.”
“We spent the last 135 years rendering people into robots. We still do that,” he states. “You cut off all interesting work and creativity from people and render them into robotic form.”
RPA, he says, “takes the robotic work out of what humans do, leaving humans free to be doing things they are much better at doing than what machines are able to do, and probably in some cases will never be able to do.”
RPA looked relatively simple, he says. “It was designed for business office people, not IT people, and gave you pretty quick wins in a relatively easy way.”
He distinguishes it from AI and cognitive automation, which he says, is “trying to put humans into the robot, to replicate what human minds can do with computers.”
“What you should be trying to do is produce machines that do things humans can't do or don't want to do.”
According to Wilcocks, the leaders in RPA deployment get a “triple win”.
These are through shareholder value (hours back to the business, workforce flexibility), customer value (remove pain points, enhanced customer journeys), and employee value (more interesting work, enhanced reputation as an innovator).
Many of these benefits are unanticipated, he says.
He cites some examples across government and private sectors.
The Sefton Council in the UK used RPA way back in 2015.
The RPA programme allowed back office processes to be handled by digital workers with 100 per cent accuracy, makes the process auditable and allows their employees to work on higher value tasks. The project released two full-time employees to focus on more strategic work.
The council started in the revenues department, signing people up to direct debit payments for council tax, indexing documents, and assigning them to specific workflows.
The council reduced the council tax direct debit payment input times by 80 per cent - with 800 debits received in three weeks completed in 19 hours, compared to 92 hours.
The cost of the transactions went down from from £1 to 20p. The transactions are processed on the day of receipt, and allows for better customer service.
A utility company in the UK met regulatory compliance deadline in six weeks using RPA that they already had. They used to require 35 new people, who took three months to do the job.
He says RPA is also thought of as back office - but there are “endless ways” customers are benefitting from it.
A major telecommunications company used RPA to reduce the time to replace the SIM card from two days to 20 minutes.
Shop Direct in the UK has lots of customers in areas that get flooded. They developed a process where they contact people in distress during the floods. They were able to delay payments that were owed for purchases, but on a one-on-one basis.
“They enhanced the customer journey”.
Not a ‘no brainer’
By 2017, he says the research showed a lot of people were running into challenges with RPA, which led him and co-author Mary C. Lacity to publish Robotic Process Automation and Risk Mitigation: A Definitive Guide.
In their studies, they identified 41 risks across the RPA lifecycle.
“The organisations are not managing these risks. They are not even identifying them, and in many cases, are assuming it is relatively risk free.”
RPA was not necessarily the “no brainer” he thought about in the first book, he discloses.
This was also confirmed by reports. One of these, by EY, reported seeing 30 to 50 per cent of initial RPA projects fail.
HFS talks about satisfaction rate of 52 per cent in march 2019 for RPA projects. “In other words, 48 per cent are not satisfied.”
McKinsey also talked about localised applications that can not scale, he notes.
He calls on organisations to be wary of “massive RPA washing by vendors who are on an exponential sales curve”. This can include overstating capabilities of RPA, overselling RPA and cognitive automation, thus causing confusion in the market.
Wilcocks explains the RPA market is split into:
The ‘leaders’ (20 per cent) who are doing extremely well;
The ‘followers’ (25 per cent) who are trying to do well but haven’t got the management capability yet or might have chosen the wrong tool for what they want to achieve;
The ‘laggards’ (35 per cent) who are not doing much at all; and
The ‘also rans’ (20 per cent) who have almost given up or are almost not running at all.
“The interesting thing is that with RPA, the killer app is not the technology. It is the human that is using the technology,” he says.
In his studies, 25 per cent of the organisations saw technology as the key problem; while 75 per cent say it is managerial and staff or the organisational structure around it.
Another interesting insight from his research was that “people were not fearful of this technology, they embrace it.”
He says one of the big black holes in all studies in automation is that they never factored in the amount of the increase of the work to be done as a result of exponential data explosion and bureaucracy.
“These are problems technology creates,” Wilcocks states.
“When people experience work intensification and there is rising tide of work to be done, they embrace technology when they see that they have got more interesting work to be done .”
“It is important to develop automatability criteria for RPA,” he adds.
“One of the tripping points is choosing the wrong process to automate,” he says.
“It has got to be stable, mature, and you are able to write rules for it.”
Another golden rule is that organisations really need to optimise the process before they automate these.
“Always test capabilities in live operations in a controlled experiment before you allow it to go wild.”
He also points out that the “real riches” to be had is doing RPA on a big scale.
“RPA is a scale game and yet most organisations have not scaled their RPA.”
There are many reasons for this, but he sees the main reason is strategically positioning RPA in the organisation.
“The result is they get quick wins, see small opportunities, and they don’t go beyond localised applications.”
According to Wilcocks, organisations can accelerate learning and get into the “triple win” phase by becoming strategic from day one.
He says the leaders see RPA as integrated with what they are doing with cognitive or digital, or the long-term business strategy.
“Because it is strategic, they have heavyweight business people involved at the front end, they fund it, resource it and protect it.”
“They don’t leave it to the IT department,” he says. “It is a business imperative that you are trying to deliver on, it rests on business operations or above.
“But ICT should be brought on board early. This is absolutely necessary because ICT provides the infrastructure for RPA,” he notes.
“Its automation is highly dependent upon ICT.”
IT can help with infrastructure configuration and scalability, as well as prevent “shadow IT” and proliferation of “RPA islands”.
Wilcocks says IT will also benefit as RPA reduces the backlog of business operations requests.
Governance is critical, he stresses.
“You have programmed governance, not just this is a one-off project and we will drive it through then go do something else.”
“Who does what? Who is responsible for what? Who is in charge of making decisions?”
The leading organisations have these sorted on day one, he states.
“No longer should you see RPA as a mere tool,” he says. “It is actually a foundation or platform in which you can increasingly plug and play cognitive technologies, as well.”
Organisations will not lose by building their internal capability to do RPA, he says.
Read more: Forum aims to reach and teach Kiwi youth about cybersecurity
RPA leaders have built internal capability using advisors and partners, he adds.
“Many management consultancies have collective experience of people who can help you accelerate your learning and build your capability as well.”
So what is missing in the current discussion on RPA?
“I still think the skills shift is the big missing issue,” he tells CIO New Zealand.
“People still get hung up about job loss,” he explains. “But if you look over history, all technologies have job loss and job gains.”
He says the PC between 1987 to 2010 created in the US 14.5 million jobs.
Wilcocks goes back to an earlier shift. In London, there were around 350,000 horses being used for transport in 1900. By 1926, these horses were gone, as well as related jobs, which he says impacted around 10 per cent of the population.
“But how many jobs were created between 1900 and 1926 that were motorcar and bus-related? The answer is massive.”
This, he believes, will be the same with RPA.
“They are creating new processes, new lines of business, changing the business model. They are all creating new jobs around that.”
Government, obviously, is in control of a lot of education and training, he says. “I think they really have to do a lot more work in this area to enable education and training in these types of skills.”
He notes that the skills are not all STEM-(science, technology, engineering, maths)-related. “In fact, a lot of the skills that are needed more than ever are distinctive humans skills that machines can’t do.”
These include social interaction, emotional intelligence, teaming, leadership delegation, and creativity.
“People underestimate the importance of education in these areas,” he states.
Government also has a role in ethics around these technologies and making sure these are utilised for good reasons and do not have bad consequences, he states.
Corporates have to get out of the habit of just buying labour off the market, he states.
“They have got to reinvest the money they earn from automation into retraining, education, and reskilling their workforce.”
“This works to their advantage,” he adds, “because otherwise they would have to buy them off the market at a higher price”.
He stresses the onus also lies on the individual worker, who should not be passive amidst the changing work landscape.
“The modern individual has to be a self-learner, they have to see where the skills are moving to,” he advises.
Wilcocks leaves with a thought from Blue Prism chief evangelist Patrick Geary:
“You have to plan for where this is going to be, not where it is now. You have to build a foundation for a tower block, not a bungalow.”
The author attended the ‘Moving to Strategic RPA in The Public Sector’ forum in Wellington as a guest of BluePrism.