We Developed The RPA Platform Digital Exchange To Free The Makers To Innovate
Blue Prism CEO Alastair Bathgate co-founded the company more than 17 and a half years ago. He was a pioneer in the robotic process automation (RPA) field before the term was coined. There are a number of aspects of his company that go against expectations. This tech leader is not based in Silicon Valley nor in the US more generally. It is a British company, but it also is not based in London. It is based in a small town in the northwest of England called Warrington. There Blue Prism has access to great technology talent from the Manchester University, which Bathgate notes is the university where the computer was born.
Bathgate notes that the value proposition for RPA has changed from cost savings, largely through cutting staff to revenue enhancement. His team has been hard at work at developing a new Digital Exchange, an RPA platform, which he believes will accomplish his ultimate mission to free makers to innovate, while positioning his company as a hub for RPA. He highlights all of the above in this interview.
(To listen to an unabridged podcast version of this interview, please click this link. This is the 33rd interview in the Tech Influencers series. To listen to past interviews with the likes of former Mexican President Vicente Fox, Sal Khan, Sebastian Thrun, Steve Case, Craig Newmark, Stewart Butterfield, and Meg Whitman, please visit this link. To read future articles in this series, please follow me on on Twitter @PeterAHigh.)
Peter High: I thought we would begin with the genesis of Blue Prism in 2001. You are the co-founder of the organization with David Moss. Could you talk about the inspiration behind the idea?
Credit: Blue Prism
Alastair Bathgate: Dave and I worked for another software vendor that did collections, recoveries, and functionality. Our first customer was a bank in the UK, Barclays. We had previous hands-on experience with centralizing collections and recoveries for checking accounts and personal loans into one call center in Manchester, and that exposed a number of issues. It created a lot of efficiency improvement. The conversation that sparked Blue Prism was from the head of change there who made it clear that there was a wide unmet automation need.
They were unable to address that with traditional IT methods. It is not that IT was obstructive, they were very helpful, but they were building the big-ticket items, and there is a certain economic granularity that comes with an IT program. We recognized that there needed to be a business-led solution to this, and this eventually became humans training software robots, as opposed to techies writing code and building systems and applications for use in the business.
As I was looking at it from the other end of the telescope, we quickly realized that business-led is not enough. That can lead to anarchy if you are not careful. You need to be in control and you need some governance around it. Banks and other organizations have regulators that are keeping an eye on things such as the data security, transactional integrity, and all these other things that need to be built in. Phase two was building in those capabilities to turn this into an enterprise-strength version of an RPA.
RPA as a term was not coined until about 2012. The term robotic came from us and it became robotic process automation. That was truly when the new category, the new market for RPA, was founded in its early days. It is certainly not mainstream now, and it is still very early in terms of its adoption curve around the world, but at least it is a recognized category now.
It feels like a long journey, but there have been some logical steps on the way. The services work that we built in version 0.1 for Barclays was their code, actually. In the end, that all went in the trash can and we started from scratch. It was a great learning exercise for us. Then we built the product, scaled, and then rolled out to market over the last six or seven years.
High: Financial services is a logical place to begin given the focus on and sanctity to information. What other industries have been the biggest adopters and why?
Bathgate: You are right. Financial services, and particularly retail banking and insurance, are probably our top two verticals. Other forms of financial services and manufacturing, as well as telcos, retail, health, and health insurance are big too. You can imagine lots of knowledge work taking place. Blue Prism itself uses its own product, and we are only a 700 person organization. You do not necessarily need to be a super large company to make good value transformational use of digital labor.
High: That is such an interesting point you make: that by using it yourself, you also have a very efficient employee base.
Bathgate: I wish it was a bit more efficient. The pains of growing 10X people in just over two years is something to consider. We try to eat our own dog food, as they say.
High: As you make the case for Blue Prism and as the executives within your client companies do the same within their organizations, what are the metrics you counsel them to track most vociferously to validate the value of RPA or Blue Prism?
Bathgate: I think that is a trend in the market. If you look five, six, or seven years ago, when RPA was just in its foundational stage, people were looking at reducing labor input costs. If I have 250 people in the office next door, and if I could replace 100 of those with 50 robots, then I am going to make a great labor savings, and that is a valid case. People started talking about reduced costs, and then people started talking about hours back to the business.
It was not always about killing jobs. In fact, we have discovered that most of our customers were not eliminating jobs at all, but they were reassigning people to more value-adding tasks. They thought of it in terms of creating capacity and giving those hours back to the business. People are even more sophisticated now and people are starting to say, “If I had digital labor 10 years ago when I built that process, I would want to build it another way. I built it because I have been constrained by resourcing it through humans rather than through digital.”
People are starting to say, “Maybe we should redesign our processes altogether to take advantage of this.” You have got something that needs to be done on the first day of the month and it takes 20 people. How do you resource that with people? I will tell you how you do it: you hire one person, and you get your report a month late. With Digital Lab, you can just throw flexibility at it, hit it on day one, and then use the digital labor for other things across the rest of the month.
It is not about cost reduction necessarily, and it is not about labor unit cost reduction. The benefits that we are seeing are coming out in increased revenues sometimes, they are coming out in improved compliance, and they are coming out in speed of service.
One of our case studies in the UK was Telefonica 02. They had a SIM swap process that took 24 hours. Customers were saying, “I bought a new working phone and now I am going to have one not working for up to 24 hours. This is crazy.” There was a fairly simple process that needed to be done to connect the new tablets and the SIM card to the network. That was done in India in an outsourced offshore center. Because it was offshore and outsourced, if there were any kind of peaks, then there was a delay. They had to put the service level around it. They automated that using Blue Prism, and it is near real time. I am a customer of 02 and last time I went, they handed me the phone and almost instantly, you saw it come to life. There was no visible difference.
There were 50 or 60 jobs at the outsourcer in India that were then applied to it because there was a stack of stuff waiting to be offshored. Again, there were no jobs eliminated but they got a new chunk of work and there was a savings. Notionally, they have created some capacity and some cost savings, but that was not the primary benefit. The primary benefit was that the customers were walking out the stores happy with the phone that worked. That was the start of people starting to think about new benefits from RPA rather than thinking solely about existing processes to be automated. The thinking is, what are we are trying to achieve and should we use digital labor, human labor, or some combination to achieve it?
One of our financial services customers had a duplicate invoice payment problem. People were presenting invoices twice and getting them paid to the tune of about $46 million a year, which is now eliminated completely because they can now get the digital workers to double check and make sure that it has all been sorted.
We have an airline that uses digital labor to monitor the space in the bellies of the passenger aircraft because they can sell that space for freight, but you cannot put a human on saying, “Peter has checked a bag, this changes the route, there was a bit more space, you can sell a bit more space.” Digital workers can do that. The salesforce now gets a near real-time report that keeps updating noting, “Space to sell.”
First, it makes it efficient. Secondly, it is better customer service because people are waiting for freight to ship, so they get it shipped quicker. Thirdly, it brings in extra revenue because flying empty space is not very efficient but equally, it is not very green either. This is the kind of stuff that people are doing nowadays. This marks a new phase in the market when people understand what digital labor is, what RPA is, and what it can achieve. They are thinking much more broadly than labor cost reduction.
High: The Chief Information Officers, Chief Technology Officers, and others who would lead an implementation of Blue Prism are on different rungs of the ladder in terms of the adoption of RPA. For those who are either just dipping their toe in the water or have not yet, what are some of the precursors to a great implementation of Blue Prism or of RPA in general?
Bathgate: There are still a ton of customers in the market who have only dipped their toes in the water, including a lot of our customer base. An important piece of our process architecture is to help customers get up the adoption path. What we discovered is that use of the Robotic Operating Model, which is our methodology and which sits alongside the product itself, is probably the single biggest determinant of success or failure.
The general principles are around putting a structure behind how you identify processes, how you build them, how you develop them, and how you automate them. You must remember that this is business-led technology, but if you treat it as a pure business project, then you will end up out of control. At the same time, it is not a pure technology project. If you treat it as a technology project, you will end up with too much governance, too much control, and heavy weight stuff getting in the way. Then it does not get done. You have to find this middle ground. The robotic operating model was a way of finding a new way of delivering business-led but IT controlled projects.
Our Chief Customer Officer is starting to roll out what we call the success accelerator packages. We go to the market indirectly, and all the delivery is done by our partners. Most of the big consulting firms and some very successful niche specialists are delivering great stuff on behalf of the customers. We try to make sure this is our best practice so we can ensure the customer has the best chances of success.
We have discovered that by making this intervention, the customer can get up to eight times more success in terms of their roll out. Obviously, that comes back to the partner in terms of services, or it comes back to us in terms of license fee, so everybody wins. Scaling up is the big issue. We feel like that is one of the differences between Blue Prism and the competition. Blue Prism’s scale and our connected RPA vision is all about building a capability across the whole enterprise.
High: You have started to talk about the ecosystem that you built, including the consultancies and integrators who are involved in implementing Blue Prism with various companies. How have you thought about the broader ecosystem in which you play? Where do your products begin and end?
Bathgate: That is a great point because the ecosystem, as you rightly say, was a go-to-market distribution strategy for us. We have something called the Digital Exchange, the DX, which is going to be the center of our universe for years to come. The DX turns Blue Prism into a platform. It enables everyone from delivery partners or Blue Prism employees to publish assets on the digital exchange that our customers can then download. They appear in their Blue Prism environment and then it becomes drag and drop. As an example of what you can do on the platform, you can go to Microsoft Azure and say, “Hey, what does this document say? It has got handwriting on it. I am a digital worker but I cannot read the handwriting.” Microsoft has some technology in the cloud that can say, “It is a loan application.”
The traditional way of doing that would be to use Blue Prism, Spy Tools, Wizards, Page Through, and through Azure to where you get that and enable it to open and then catch that. If that ever changed, you would have to revise that process. It took a lot of time to build it the first time around. Now, that skill is just dragged and dropped into your process flow, you just configure to say, loan application, or whatever the configuration is, and it is there and running.
In the next version of the product, we are going to bring in from the websites the ability to do that through the product itself. Blue Prism becomes a genuine platform that can access a massive range of other AI or other business applications.
The other exciting thing about the digital exchange is the ability to have private assets on the cloud, which means our customers can publish something if they wish to share it with the community. If they want to create something that is just for them, then they can keep it private. They can create an entire configuration library of every asset that they ever need, including public ones from 3rd parties and ones they can buy as well. It is becoming drag and drop integrations and automation. We are fulfilling the promise of being business led, but equally because it is bolted together and pre-API and pre-controlled and authorized, it is in control as well and you know it is going to work.
High: Can you talk about as a UK-based company, the talent market in the UK? I am curious how you perceive the UK versus the US for instance, in terms of technology and talent.
Bathgate: A large number of our people are Sales and Marketing people, but we have technical people in various locations. We recently set up Blue Prism Labs, and we chose to locate that in London. The reason we did that is because we thought the top scientist, the Ph.D. research guys, are coming out of companies like Deep Mind and the like out of London, and we went there for talent.
Our HQ is in Warrington, a small town northwest of England. We are close to Manchester University where the computer was invented. We have had access to a pretty good talent pool here, but we are looking to build our technical capabilities in other parts of the world as well. The valley is hard because salaries are so crazy, but there are many other innovation ecosystems. Austin, Texas is our HQ in the US, for example, and there is a wealth of talent down there in the software industry. There is more talent available there at a lower cost. You do not want to compete with Google for talent because it becomes difficult to get the A-players.
High: Where do you see the future of the organization. And how do you see RPA continuing to evolve?
Bathgate: Blue Prism has consistently been the thought leader. The Digital Exchange was our idea. The other guys have got some version of that. Our connected RPA is unique to us, and we think that is what turns RPA into the transformational enterprise play. The market itself may identify other things that you can do with RPA. Desktop productivity is where some of the competition are going. We are trying to halve the number of desktops rather than being a desktop system, and we are trying to create capacity in a bulk scale kind of way.
There is room for a number of different vendors in the market. Some of them have different plays in terms of their degrees of Artificial Intelligence or capability of dealing with non-standard or unstructured data and information. I think we will see a proliferation of different plays on RPA. Our intent is we want to be the platform, and we want to be at the center of the ecosystem. That is why the digital exchange is important to us. We want to be the operating system that powers the digital enterprise.
High: You mentioned AI and machine learning, both of which are related to RPA and its broader ecosystem. As the evolution of RPA and the evolution of ML and AI continues to expand, how do you see those interacting?
Bathgate: Another great question. The digital exchange is a way of us to partner with stuff that has already been built. There is no point in us, for example, committing R&D dollars to try and outcompete Microsoft in certain categories that they have already built. We are not going to build a chatbot inside Blue Prism, just by way of an example. That will always be a partner opportunity for us.
That does not mean that we are not doing our own AI development internally. At Blue Prism World London, we announced a product called Decipher. The point of Decipher is to give the digital workers eyes in order to give them insight into their financial costs with a picture of an invoice. How do you read that? It is not digital. We have to use some Optical Character Recognition (OCR) technology, but it is not just about recognition, it is about understanding as well.
What Decipher does is it can look at an invoice and retrieve the invoice number, for example, without even knowing where it is. Because it will look for the strings as it is an invoice number, it will then retrieve the number that is next to that. That gives it a good shot at getting the invoice number right. If it cannot find the string invoice number, but it can find a reference number and is looking for the invoice number, then it may flush that out to the employee and say, “Are these the same things? Does the reference number mean invoice number?” The human can look at that and say, “It does.” “All right, I will not ask again. I now know. Next time I go get one of these documents, then that is what it means.”
This is also about machine learning. Machine learning is not just about learning from humans, it is about learning from systems as well. This is another trend in the industry that we are moving towards. We are moving away from humans teaching robots how to do logical processes to digital workers. That is why that terminology needs to move on from robots. Digital workers learn either from humans or from systems themselves. That is also moving us away from semi-structured data towards completely unstructured data. Imagine things like legal contracts, for example, where key clauses can be quickly identified.
High: As a result of machines teaching machines, for example, and digital workers who can work better and more efficiently than humans can in repeatable tasks, how do you see that changing roles or skills of real workers? Which ones do you think will be on the rise in importance? What should the 18-year old study at university?
Bathgate: All the soft skills: empathy, relationships. One of the things I say in my keynote is with video conferencing invented 20, 25 years ago, why do we ever get together in person? The answer is because there is a human connection. People like to see, feel, and touch. There is an emotional connection. I do not think that is going to ever be replaced by technology. Building relationships is a key thing for humans, but there is also a variety of work. Someone did an analysis of all the different job titles that you could possibly have and the percentage of your job that could be automated. I am pretty okay because apparently, only 6% of a CEO’s job can actually be automated using digital labor.
If you think about another case study that I saw recently at S&P in New York, they have a bunch of very clever and very well paid financial journalists whose job it is to go out and write insightful stuff on companies. They were precluded from doing so by having to write stock reports. Very mechanical stuff. They got Blue Prism, Appian, and Automated Insight. They use that combination of technologies with our Digital Exchange platform to automate the production of stock reports. The benefit is not about saving costs, it is about freeing people to do what they are really supposed to do, which is to provide insight.
Again, I do not think technologies are capable of that yet. If you are a journalist who provides insight, who writes in the monthly reports, then you have still got plenty of work to do. If you are just working on reports: the news, sports reports, just the very factual basic stuff, then hopefully, you are going to be relieved from that stuff fairly quickly if you have not been already.
High: Can you talk a bit about connected RPA?
Bathgate: Connected RPA just completes our vision of being the enterprise play here. A lot of people check their phones in the morning before they even get out of bed. And then you get into the office and you are handcuffed and constrained by the systems and governance of your employers. If you are an innovator, if you are a maker, you are going to get a bit frustrated. That is why lots of little spin-offs and tech startups start to nibble away at the very fabric of the profitability of the business. This is particularly acute in FinTech with banking, for example. What Blue Prism tries to connect together is these makers in the organization: connect them to the company strategy and connect them together, because you have people in your organization who have the ideas, but they cannot implement them.
It is a bit like the original Barclays tale. You have 250 items and I know what I need to do, I just cannot get it done. I cannot build that level of operational agility to enable me to react to these things that are happening outside of the organization. Instilling the organization with that level of operational agility is an important part of connected RPA. Connected RPA is not just about connecting people together, as I mentioned before. It is about connecting the technologies together, as well. The digital workforce in its wider sense is something that becomes part of the connected RPA vision, and we ended up with a connected entrepreneur enterprise.
I do not know that entrepreneur is the right phrase actually, but certainly innovator and maker: the people who have great ideas but are constrained by the governance and systems in your organization. Freeing the makers to innovate keep your company relevant to the competitive world.
High: Is there a certain skill that is necessary to implement Blue Prism to work for Blue Prism?
Bathgate: We are a 700 person organization and we are finding uses for tons of digital workers, Blue Prism digital workers. A digital worker for us is a 24/7 available resource. It can do the work of at least two people. That is pretty productive. If we can do that, there is a very, very large number of organizations in the world that can benefit from digital labor.
Peter High is President of Metis Strategy, a business and IT advisory firm. His latest book is Implementing World Class IT Strategy. He is also the author of World Class IT: Why Businesses Succeed When IT Triumphs. Peter moderates the Technovation podcast series. He speaks at conferences around the world. Follow him on Twitter @PeterAHigh.