by Peter Bendor-Samuel

A fundamental flaw in approaching digital transformation

May 16, 2017
CIOCloud StorageTechnology Industry

Just implementing digital technologies and learning how to use them won’t create a big performance impact.

What makes digital technologies so different and disruptive is their potential to enable very substantial business benefits. “Enable” is the key word. Too often, executives see the power of a technology and reason to themselves, “This technology will create significant benefits, so we need to implement it and learn how to use it to our advantage.” The problem is this is a fundamental flaw in approach that almost always ends up in a digital transformation failure.

Technology does not drive change; creating substantial business value requires changing the business model. And a business model change requires many changes in operations, not just new technology. Yes, those changes are cross-functional, usually end to end, and always disruptive. But without changing other operational aspects than technology, the transformation initiative will fail to deliver the anticipated outcome.

Take the case of robotic process automation (RPA), for instance. It’s the simplest of digital disruptions. The technology is 30 years old, and often it’s no more complicated than screen scraping or algorithms. It’s a very simple concept, and it looks so easy. Yet there are very few cases of companies with more than 100 bots installed. Why is that? It’s because they took the approach of thinking they would implement the technology, learn how to use it and it would create value. Well, it doesn’t happen like that.

Yes, there are cases where companies implement RPA a few times throughout their organization. But then they run out of steam. That’s because their plans didn’t take into consideration the other changes that need to happen, which are much deeper and much more difficult than the implementation of the technology.

For example, the first thing you find out when you implement RPA is your OCR (optical reader) isn’t of sufficient quality to create reliable data. It’s good enough for humans but not good enough for a robot. But there is no budget set aside to replace the OCR technology. And the OCR technology often sits in a different department, which has no interest in changing the technology and has no budget to change it. So, nothing happens. As Peter Drucker advised, “If you want something new, you have to stop doing something old.”

Besides the OCR issue, companies encounter security issues when they implement RPA. The security clearance process is designed around people, not software robots. The RPA robot must face the same security clearance that a person does, but most policies and processes are not set up to accommodate that. Therefore, the company must rework its security protocols (again, involving another department with very little interest in doing it). The result: passive-aggressive behaviors that slow down the progress or potentially stop the project from scaling.

Another issue arises when applying RPA necessitates rethinking the business process. As you start to automate functions or tasks, the underlying process needs to be restructured. But, again, there is no budget for that or no capacity to drive that change. Without restructuring the process, you can’t drive the automation deeper.

Non-reliable data due to the legacy OCR, robot security clearance issues, rethinking the underlying business process — it’s no wonder companies run out of steam and don’t achieve the business benefits they anticipate from applying RPA. And it’s the simplest of digital technologies.

Need another example? Consider what happens when moving to the cloud. In theory, moving work into a pay-as-you-go infrastructure world is cheaper than having dedicated servers with low utilization. In reality, moving a few workloads over to the cloud increases your cost because now you have the cloud cost as well as the cost of the existing environment. You have security and resilience questions around how you’ll operate in the cloud. There are management issues too. A lot of rethinking underlying processes.

The change required to adopt a digital technology (whether the change is policy and process or other investments) is well beyond the benefit of the initial savings. People throw up their hands, and ambitious agendas get defeated.

Having said that, the reality is digital technologies are potentially very powerful in disruption — creating new value and competitive differentiation. But just buying a digital technology and learning how to use it won’t deliver real business benefit. If you want the benefit, you must recognize that there is much more change associated with it.

My advice: don’t drive your desired change from a technology-first perspective. Instead, drive it from a business-impact-first perspective. Make sure your approach to transformation is not about the technology or the disruption. First, understand what the disruption enables, then look at the business benefit and then drive it from the business benefit back to the technology.