Change Your Mindset to Avoid Digital Taylorism

When one of my esteemed colleagues was trying to explain for the umpteenth time in a client meeting how Arago’s approach is different from RPA, his way of explaining it triggered a whole stream of consciousness on my part. To underline that RPA is essentially ringfencing (if not embalming) industrial processes, i.e. automating at the end of highly repetitive processes just like a physical robot would do in an assembly line, he made the compelling point that Arago neither uses or needs scripts or decision trees for its Agile Automation. More pointedly, working with our AI platform does not require that processes be re-engineered into a series of steps that can then be recorded as scripts or bots that are managed in decision trees and triggered by human assistance.The stream of consciousness had little to do with the actual technical details but more with the fundamental question of what outcomes are we trying to achieve with automation and AI?


Judging from those discussions and reflecting the current industry commentary, we seem to have lost track of the initial goal of process automation that was meant to get us closer to the notion of straight-through-processing that could accelerate the journey toward the digital process nirvana. Rather, it feels to me that the industry has lost its footing and is running in circles as it skids on the Kool-Aid splashed around by some of the RPA marketing. But most of those RPA deployments are task automation with attended bots. In other words, clients are using them to integrate across disparate, if not piecemeal, technologies. The legend that is Mike Hobday (erstwhile IBM’s European Automation supremo, now in a similar position at AntWorks) was trying to make a similar, albeit much more colorful, point that RPA is in danger of becoming a Digital Frankenstein given the static and repetitive nature of the technology. While Mike was mischievously trying to promote the integrated automation platform of his new employer, the lively ensuing discussion rightly pointed out that automation exists because the process design is either flawed or legacy. Before folks cry out that the integration of AI capabilities will change RPA from such a Digital Frankenstein into a Digital Superhero, consider the following lucid point. In a different context Ron Schmelzer, analyst with Cognilytica, suggested: “Many firms are claiming to be AI-enabled when all they have done is put some thin capability provided by a third-party library or API that doesn’t really transform their existing product into something that is inherently different with that new intelligent technology. One of the biggest offenders of this AI-as-a-buzzword is the entire RPA market.” The point I am trying to make is not RPA bashing, but recalibrating the discussions more toward the outcomes. Clearly, these arguments that I have highlighted provide food for thought. So, what can we learn from those deliberations for guiding clients?


If we leave aside the marketing noise for a moment, for me one of the key issues that clients are struggling with is the lack of clarity for the direction of travel in order to drive change through the organization. And we are seeing that in our discussions with clients and prospects. Often, we get buy-in from the C-Suite, with paranoia about disruption being one of the key drivers. However, the moment the lead passes down to the operational teams, those discussions often stutter. And this is because the operational teams are seeking more “traditional” outcomes such as cost, efficiency, and talent. All this while working within the bounds of “business as usual” fitting a process to the needs of this year’s latest automation software tool.

A recent survey by Accenture provides more color on those challenges. The data suggests that 84% of C-Suite executives believe they won’t achieve their growth objectives and strategic priorities unless they deploy AI. Yet, 76% of those executives acknowledge they struggle when it comes to scaling AI across the business. Even more pertinently, 3 out of 4 executives believe that if they don’t scale AI over the next 5 years, they risk going out of business entirely. For the authors of the study, a Proof of Concepts “Factory” (I have heard much more colorful descriptions of the same phenomenon) is one of the key reasons for the lack of scale. This “factory” is characterized by the following characteristics:

  • Analytics buried deep and not a CEO focus
  • Siloed operating model typically IT-led
  • Unable to extract value from their data
  • Struggle to scale due to unrealistic expectations on time required
  • Significant underinvestment, yielding low returns

Among the levers that Accenture suggests would help to strategically scale deployments, two stand out for me. First, an experimental mindset that helps to achieve scale and returns. Second, a catch up on digital/AI/data asset debt. Thus, we are back to the discussion on avoiding creating a Digital Frankenstein.

In whatever shape or form we define a Digital Mindset, be it customer-centricity, collaboration, networks, DevOps or something else, the temptation to embalm industrial, repetitive processes with task automation won’t really get us much closer to that mindset. On the contrary, it reminds me of the quip about ERP: it’s like concrete in a basement, thus you won’t be able to move home. Ringfencing industrial processes might not feel as heavy, but it equally will prevent you from transforming. We all love to point to start-ups like Lemonade or Revolut as digital benchmarks, yet we have to be cognizant (and honest) that they don’t have to deal with technical debt. Thus, we are back to the necessary mindset that allows organizations to bridge both worlds. That mindset must include providing partners access to production data when undertaking proofs-of-concept. Only then can partners deal with the real problems but it also confirms that executives are actually serious about driving change.


We at Arago are at times guilty of leaning too much on technology innovation and the direction of travel. Be it emphasizing the knowledge economy, capturing knowledge instead “just” data or having developed new classes of algorithms that generations of students are likely to study. But the capability that helps clients to bridge the industrial and the digital mindset is the fact that our HIRO™ platform allows clients to progress to Agile Automation without having to change their processes. They don’t have to standardize or even to transform their processes to get much higher automation rates and agility that deserves the “digital” moniker.

The main reason for that loops us back to the introduction of this blog because we don’t have to engineer a static process that must then be captured in bots, scripts or decision-trees. This provides not only the agility clients are trying to achieve, but also significant cost savings as maintaining scripts and decision trees is one of the biggest costs for clients given that every real-world change to the process requires some process engineering to define a new static process to update the script.

Yet, not least given the unavoidable marketing hype around all things automation and AI, clients really have to reset their mindset and must be willing to think and act beyond what they have heard before. In a nutshell we automate the inherent knowledge of folks who manage digital operations to complete the process and not the steps. By definition we view processes as constantly changing and we use AI to easily adapt to it with minimal fuss. It is essentially like how a kid is learning stuff. Whereas the industrial mindset is more akin to having kids listen to endless tapes of instructions given to their parents’ generation. On a more sober note such an adaptive mindset can help to bridge the different expectations of the C-Suite and the operational teams. The adaptability makes processes future-proof while yielding significant cost and efficiency gains at the same time.


Perhaps we should heed the advice of some of our advisors to be a bit bolder and louder. Perhaps we should shout from the rooftops: “No scripts. No decision trees. Transformation without standardization”!! On a slightly more serious note, what need to do as an industry is to recalibrate the discussions around the outcomes of Intelligent Automation. The noisy RPA marketing is surely not helpful in that respect. To progress with that I would love to hear from you, how we can jointly help the C-Suite to draw up actionable mandates that will move the needle on their digital journey. Otherwise, many organizations will end up stuck at Digital Taylorism.

Tom Reuner

Tom Reuner

Head of Strategy