Whatever happened to Intelligent Automation?
The fact that UiPath’s refreshingly humble CEO Daniel Dines made it to the cover of Forbes as the first “Bot Billionaire” tells many stories. Top of my mind are memories of when I met Daniel at the first-ever Robotic Process Automation (RPA) conference in New York back in 2014. Back then we both had no idea (though I am sure Daniel was much more optimistic than I) that RPA would be one of the biggest topics in the IT industry. Fast forward 5 years, multi-billion-dollar evaluations for the three leading RPA providers are not only making the headlines but also contorting the discussions on automating processes. So much so that RPA providers are now being rebranded as AI companies. Yet, back in the days in 2014, the discussions on RPA were part of a broader discourse on Intelligent Automation. We were expecting a convergence of IT and business requirements as well as a convergence of technology capabilities. To help clients on their digital journey, collapsing the many siloes was top of the agenda. The end goal was meant to be process automation and thus enabling clients to run their processes flexibly and in a manner conducive to withstanding the digital headwinds. Yet, the current discussions on RPA appear to be confining the goals more and more to task automation. Are clients just happy to optimize rather than transform processes? Are they willing and capable to manage a piecemeal of point-solutions? Or do we need fresh approaches that take us closer to those goals that Intelligent Automation was meant to deliver? With that in mind, it is about time to revisit the discussions and take stock.
It is all about the mindset if you want to pivot toward the knowledge economy
Spending two days at the HFS Research conference in New York provided Chris Boos and me with the opportunity to get a reality check as to where the industry is at and what drives the strategic thinking of clients. Many speakers agreed that technology is just around 10% of the effort to progress toward whatever moniker we want to put on as the ultimate goal of clients’ journey. Chris put this poignantly: “To the Valley, technology is a religion, to everybody else, it is just a means to an end. Thus, we have to pivot the discussion on automation and AI back to outcomes.” Consequently, we need to think more about the mindset as well as the ambition and ability to drive change. In that context one of the most thought-provoking data points that we have uncovered from data provided by HFS is that clients are accelerating the journey to digital and are willing to either skip or reduce labor arbitrage as part of outsourcing deals. While we should read that as intent rather than a reflection of current market dynamics, it still poses a fundamental question. Namely, how can we support clients transitioning from an industrial mindset to a knowledge economy? In its current form RPA is very effective in optimizing industrial processes and replacing labor. Yet, to progress to the knowledge economy, we have to move from prescriptive automation reflecting the industrial mindset to dynamic decision-making. In other words, we have to go beyond just sprinkling Machine Learning across digital operations and discuss approaches that support making processes adaptive to changing environments and dealing with the data challenges brought on by accelerating digitization. The transition to the knowledge economy will be hugely disruptive and uneven. Discussions on automation must reflect that.
Intelligent Automation has to be something bigger than RPA
As so often, the sage of all things sourcing and automation; Lee Coulter (who has just launched his latest venture transformAI), was the sound of reason: “Intelligent Automation has to be something bigger than RPA.” What this something bigger is, though, remains open for debate. When my old chums at HFS started talking earlier this year about Integrated Automation, I had hopes that we would go back to the early discussions on Intelligent Automation with a much more holistic mindset. As the analysts rightfully stated, RPA deployments continue to struggle to scale and most IA initiatives lack an enterprise-wide strategy. But rather than including broader approaches such as Autonomics or Test Automation and pursuing a more holistic and enterprise-wide approach, the discussions appear to go backwards again. Counting bots as a proxy for automation maturity indicates that the market has shifted more toward what used to be known as Robotic Desktop Automation (RDA) – largely front-office-centric attended automation. If that is the case, how do we reconcile this heterogeneous and more employee-centric automation with the endeavor to progress toward the knowledge economy?
One of the speakers summarized the current state of RPA very succinctly. “The problem with RPA is that there are a lot of Proof of Concepts (PoCs) and use cases that haven’t scaled. A lot of enterprises find that RPA adds another step to the process and doesn’t transform it. People assumed that RPA would automatically change everything in their processes without investment in modernizing underlying systems -and that’s turning out not to be the case.” If that assessment is correct, how do we progress from here? Another data point from HFS might offer a clue: They suggest that “AI” will in 2020 surpass RPA as an investment focus. Given that “AI” currently is probably the most hyped and overused term, one way of reading that is: 1. an acknowledgment that traditional RPA approaches continue to struggle to scale and 2. the mindset is focused on integrating and managing point solutions. With OCR and process discovery top of mind, RPA and with it Intelligent Automation are changing fast. But we need a much more nuanced and outcome-focused discussion in order to help clients to progress toward the knowledge economy.
Innovation mindset and change management provide the real demarcation for automation and AI
The current state of the market might be best summarized by two seemingly contradicting statements. Many speakers emphasized that the biggest frustration of the C-Suite is lack of agility. Yet, at the same time, they remain risk-averse and inch forward with point solutions. Chris hit the nail on its head by observing: “AI requires a shift in mindset driven from the top. Otherwise we remain stuck in siloes and drown in a piecemeal of point solutions.” And one can extend “AI” as a broader proxy for innovation. He added a warning that applies to both RPA and the broader market alike: “blindly focusing on Machine Learning models is slowing down organizations. Rather, we need to solve the problems that are holding those organizations back”. In his inimitable way, Chris also reminded the audience what the broader point of reference for all those discussions should be: “Doing what customers want has done little for their bottom-line. The AI they ask for is so yesterday. We have to move to help them to survive the platform economy.” These deliberations are closing the loop as to how we should progress with the discussions on automation. We have to move from prescriptive automation reflecting the industrial mindset to dynamic decision-making in order to effectively transitioning toward the knowledge economy.
Bottom-line: Surviving Digital Darwinism requires an innovation mindset as well as the ability to drive change through the organizations
The pace of change is nothing short of astounding. To stay competitive if not relevant, decision-makers must find the optimal blend of short-term optimization and more mid to long-term innovation and transformation. The companies that will not only survive Digital Darwinism but stay competitive will be the ones led by executives who have the imagination to harness innovative technologies for new business models. However, the even bigger challenge is to drive that innovative mindset through organizations that often have battle scars from modestly successful long-term transformation projects. This is the direction of travel that the discussions on automation need to take.
Tom Reuner, Head of Strategy