A Three-Part Conversation With Professor Leslie Willcocks, co-author of ‘Becoming Strategic With Robotic Process Automation’

Do you see any partnerships forming in the RPA space?

This is going to accelerate, and, yes, there are quite a lot. Blue Prism, for example, have a strong partnership with the company Abbyy, that traditionally did OCR (optical character recognition) and they realised that that was something that would fit at the front end of Blue Prism because they needed to get more structured data in order to use the Blue Prism RPA tool. So Abbyy did that – added optical character recognition. If you look at a company called Thoughtonomy, that again was a Blue Prism allied company that also has OCR technology at the front end to take unstructured data and convert it into a form that a Blue Prism tool can take. And only in July 2019 in fact, Blue Prism bought Thoughtonomy. So you might see these sort of complementary technology companies starting to buy each other. WorkFusion always had a cognitive technology and actually had to develop an RPA capability, in-house, because it was too far ahead of the market with its cognitive technologies. Uipath bought a cognitive technology company several years ago, in order to enhance what it could do with its robotic process automation tools. Increasingly, these companies – the top three especially: Uipath, Blue Prism and Automation Anywhere - are trying to develop integrated automation platforms whereby you can drop the cognitive technologies of their partner companies into the platform. You can then utilise them quite quickly in a complementary way with RPA.

So that complementarity has become the key to moving forward this year. And you will see more mergers and acquisitions between the RPA and the cognitive technology companies as we go forward. But it’s more complex than that. Let me see if I can describe it to you, because a lot of people would not have this model in their head: You see, there’s a sort of pincer movement going on towards creating these integrated automation platforms. There are the RPA providers, who are either partnering with cognitive technology companies or buying them up or developing in-house cognitive technologies. Then there are the sorts of ‘big players’, what I would call the systems integrators that are looking to orchestrate RPA with cognitive analytics and AI – people like Accenture, Wipro, TCS, KPMG, and GenPact. They all have orchestration platforms to try and bring these things together. Then the other movement that has already begun, are the companies that produce cognitive or analytic products. They now begin to see RPA as complementary and are sort of going back into RPA suppliers – companies like IPSoft, Arago and Haystack. You then have the more long-term picture, which is enterprise software producers like SAP with their Leonardo platform; SalesForce with Einstein; Oracle with Adaptive Labs; and Pega with their OpenSpan platform. They are trying to integrate RPA with cognitive and artificial intelligence tools, trying to integrate those into their platforms. So there’s going to be a lot of movement in terms of mergers and acquisitions across those four sets of players.

Can you comment on the positioning of the top players in the RPA space?

If you look at Blue Prism, they go back to 2003, and created a fairly robust RPA tool to work with a German utility which was a pretty difficult, demanding client. So they developed quite a reliable enterprise-safe tool back in 2005. UiPath was working with Microsoft, Google and IBM developing self-development kits for their different service offerings back in 2010 & 2012, and then realised this market for RPA. They didn’t get it right initially, worked with an Indian BPO service provider, and came to understand the market, then got RPA. They developed a more open source-based technology piece of software, which often appeals to a different group of people than does the Blue Prism tool. Automation Anywhere was working with desktop RPA to give people freedom from IT departments with directly applicable software on the desktop into the business unit, back into probably 2008.

Subsequently these three, especially, had broken away from all the others and got massive increases in funding, to the point where I mean UiPath now says it’s worth seven billion dollars and Automation Anywhere about $3.86 billion. Blue Prism’s pound shares when first released, are now worth anything between £14 and £24, depending upon the day you’re in the market – but it’s still quite a lot in terms of share value relative to what it was on the day it released its shares.

What about the technology, their cognitive automation capabilities, current progress, pricing models?

Yes, well, it’s very confusing because it’s all over-marketed. The interesting ones are the top three, as they’re just outpacing the others now because they can fund the marketing, they can fund the salesmanship, but they also can fund the research and development. And they need to move on several different fronts, each of them in a different way.

So if we start with say, Uipath. Uipath is based on an open source coding and, say three years ago, was regarded as cheap and clunky, by which people I interviewed meant that they liked it because it was cheap. They liked it because the service was responsive – both pre-sales and after sales – and it was seen as clunky because they had to do a lot, especially if they wanted to scale it to enterprise level or within a function. But generally speaking, the people who choose Uipath, tend to be technologists and engineers, who like something to do, and make their own imprint on the source code and the software. So there’s a sort of marriage here with UiPath clients and what they do. UiPath spend, and have quite a large group of people there in R&D. I would say that two years ago they would have been behind Blue Prism, who was the market leader on the superiority of the technology.

The interesting thing is that RPA, the actual software, is quite dumb. It doesn’t do an awful lot, but if you surround it by pre-packaged smart software, you can get it to integrate in reliable, secure ways with other enterprise systems, leave an audit trail and automate and store reusable objects very easily. But you have to do more work at the front end with Blue Prism, not an awful lot, but if you want to deploy it to get it right, you do more work at the front end. With Uipath, you do less work at the front end but you have to start doing a lot more when you start scaling. Automation Anywhere is a different product historically. It’s more a desktop-placed automation using a recorder bot. And the idea was that you would get much more productivity out of a single worker. It’s a sort of assisted, desktop robotic automation achieved by getting them to automate what they normally do that is physical or repetitive – tasks that needs to be done fast. Getting that to operate at integrated enterprise level safely, with other IT systems, has, historically, been a problem. Obviously, Automation Anywhere and Uipath have thrown a lot of R&D at that problem, so they would now claim to have automation platforms that could integrate, not only with other IT systems, but also with cognitive technologies. Whereas historically Blue Prism has tended to be the choice of the risk averse, highly regulated businesses, that want the thing to work unfailingly, don’t like risk, want to scale to enterprise level and don’t necessarily want to declare a UDI from the IT department, but nevertheless, want business to drive where the automation software is used, in terms of what the business imperative is.

Where does cloud based automation as a service fit in?

In our books, we talk about there being four archetypes for RPA: desktop, enterprise, self-development kits and Cloud. There are already several companies that make a business out of doing Cloud-ready RPA. So, in one case, a bank spins off a Cloud-based RPA platform for its customers to buy into. It develops the capability in-house, and then offers it as an additional service to its customers. And it works extremely well. Then there’s a company called Digital Workforce in Finland, that offers, globally, the ability to use RPA or digital workers as they like to call them, where you’ve got peaks and troughs in your demand for processing data or invoices, or whatever it is in your business. They organise that very well so that you can plug-in quite quickly and get the digital workforce on your case – within hours in some cases. So that’s definitely a movement. As always with Cloud, there are question marks about security and privacy. I mean some banks and utilities tell me we can’t go down that route unless we completely control the Cloud connection so it would be a private Cloud at best. But obviously, with other businesses, you could see it being quite a commodified service and becoming more sophisticated as time goes by, whereby people just utilise the power of Cloud-ready digital workers and just pull them down on a pay-by-use basis. This is a reminder of the old model of what we used to call ‘application service provision’.