Throughout my career, I have approached business transformation via the mantra “process follows strategy, structure follows process”. As far as I can remember, the preaching from management gurus was: “First, define the business strategy. Then plan, implement and improve business processes to assure the effective and efficient delivery of products & services”. In other words, perfection comes via the mastery of repetition. Business success hinges on defining the optimal sequence of streamlined activities that deliver the goods.
With processes, perfection comes via mastery of repetition.
Will this approach still be as relevant in the future as it has been in the past? I’m not so sure. Let me explain why.
My latest area of focus has been the transformation of the support/assurance processes for a major company. The ultimate aim: instead of solving failures (incidents) when the customer calls, failures are avoided altogether by detecting situations leading to failures and proactively solving them before anything bad happens. As you can image, there are plenty of opportunities to implement machine learning analytics in this context. After many meetings with many specialists such as process engineers, business and data architects, data scientists and many other clever people, I truly see the possibility that the “process approach” may be on the way out.
OMG, where did the process go?
How could we possibly manage a company without processes? This is how it can happen.
The first stage towards a “process light” vision is to push the key decision-making within processes to machine learning algorithms. If x and y occurs, then do activity z. The benefits of doing this are huge: once implemented, you can basically keep on improving the decision-making over time, and thus the entire process, without having to change your front-ends, back-ends, activities, roles etc.
In a second stage, activities are broken down in ever smaller parts. The learning system is expanded to consider ever more data (relevant customer data, business interactions, diagnostic data, historical data etc.). At this point, the sourcing of data and building of analytics become ever more important. That’s where the energy and resources are invested. The result is an ever greater number of “process variants”. Each customer case is being handled differently to obtain the best possible result for the particular customer and circumstance.
This transformation, pushed to the extreme, can lead to a third stage where pre-defined activities don’t matter any more. Quite the contrary, in this new data driven decision-making context, each decision and task is tailored to a specific customer situation. The enterprise is freed from the constraints of traditional processes. Business efficiency and effectiveness isn’t driven by the optimization of tasks any more, but from the optimization of data sourcing and analytics.
In a data driven approach, business success is driven by mastery of data sourcing and analytics.
People working in the organization are not linked by common workflows and systems, but receive their instructions in a timely, case-by-case basis. Imagine employees simply receiving instructions via a chatbot on their PC, mobile phone, augmented reality devices or earphones! Processes and workflow systems disappear, replaced by a central big data & machine learning cluster. People and systems don’t make decisions – or guide the customer experience – based on “what has been defined in the process” but simply implement the instructions coming from the “central brain”.
The slow decline of the business process
Does this sound crazy and farfetched? My answer to that is yes and no.
On the “yes” side, I can testify to very concrete projects and results for the first two stages of the above described transformation. I can also attest to serious attempts being done to implement the third stage. Seeing is believing as they say.
On the “no” side, I can see barriers that slow down this transformation or make it uneconomical in certain cases. Setting up a big data cluster and a robust team of data scientists is a major undertaking that requires investments with many 000’s and over a long time. Furthermore, on the business side of things, creating new knowledge and integrating this knowledge into a customer experience is difficult and doesn’t happen overnight. It’s the result of a “learning journey” that is fraught with trial and error and incremental improvements. Another barrier is that replacing activity based processes by data driven instructions requires scale to work. It’s a valid approach for big companies with many customers in which some processes are repeated hundreds of times daily and for which a data trove can be collected and made available. Scale matters. The data driven approach is neither technically feasible nor economically viable for smaller organizations or units with small amounts of cases being handled every day.
My conclusion is this one. The vision of a centralized big data brain that pushes out timely “next best activity” instructions to multiple touchpoints via connected devices is definitely not farfetched any more. The transformations to get there have started in major companies. The attention and investment directed at such approaches will keep rising in the foreseeable future. This data & analytic approach to business management has the potential not only to improve processes but to replace them altogether. So be prepared to witness the slow but gradual disappearance of the business process in data driven enterprises.
The vision of a centralized big data brain that pushes out timely “next best activity” instructions to multiple touchpoints isn’t farfetched. The transformation to get there has started in many companies.
One thought on “Is This The End Of The Business Process?”
Thank you for your insight. I wonder if data driven insights can be broken into two different approaches. Those that are easy to automate and those that aren’t. Then it follows that those that aren’t still can be informed by data intensive systems. However, we still need an understanding of both from a perspective of data processes and how we get to informed decisions from them.