The main obstacle to business progress is not legacy technology, it is legacy processes.

According to many reports, legacy technology is considered a major obstacle to progress in the business of financial institutions and many other industries. But is it really?

The purpose of any technology is to support business processes (which, in turn, create products and services) and, in some cases, to replicate and reinforce them. However, no technology is capable of creating a whole new or more efficient process on its own.

Let’s consider some examples:

1. RPA – doing wrong things very quickly. RPA became very popular due to its perceived ability to increase the effectiveness and velocity of many routine operations dramatically. While implementing RPA might indeed provide some short-term benefits, it would quite probably be strategically detrimental to the business because it digitizes the existing process with all its flaws. Moreover, due to the investments made in this digitalization, it becomes virtually impossible to even initiate any process improvement activities before the planned ROI is achieved.
2. AI – garbage in, garbage out. AI, an idol of recent years, consumes all the inefficiencies embedded in the information it is fed, generalizes them, and returns them as answers to new questions, sanctified by the use of state-of-the-art technology. These answers are often accepted without criticism. To be fair, this problem has become widely acknowledged, but it has barely affected the enthusiasm of AI apologists.
3. Process mining – beautiful buttons on a shabby suit. Process mining or process discovery should, by design, help solve process-related problems. However, it is capable merely of finding technical (though not always obvious) issues like duplicate operations and loops (and only in fully digitized processes). It has nothing to do with addressing the essence of the process. If process mining is ‘reinforced’ by AI tools, we encounter the same problem of AI being trained on bad data, leading to poor advice.
4. Large enterprise systems – casting your pain in bronze. The problem becomes even bigger with large-scale technologies like ERP or CRM systems. It is common practice to adopt “out of the box” processes embedded in the software during the implementation project. Although these processes are often presented as optimized, they are, in fact, designed to make the software work in the most efficient way—benefiting the software, not the business that uses it.

Adapting large ERP and CRM systems to a company’s actual processes consumes so much time and money that no senior manager would even discuss the possibility of changing anything after implementation is completed (or more often, when it halts) and before investments are recouped. If the system is set up to leverage existing processes, all their flaws might become irreparable for years.

Microservice architecture reduces the scale of this problem to a certain degree but does not eliminate it. It merely replaces one huge, digitized tangle of problems with several separate ones. However, the necessity to address process and product issues first, and digitize only optimized ones, remains.

What are the main requirements for a business process analysis and improvement project to be successful?

1. Business is of the people, by the people, for the people. Business processes are first and foremost interactions between people, not systems. Everyone responsible for the processes in question, including key process steps, should participate and be actively involved.
2. You never know where you'll find it, where you'll lose it. You never know which step of the business process will turn out to be a bottleneck. Otherwise, this problem would have been addressed already. Be ready to question and challenge everything, even those parts of the business scheme that seem obvious—especially those parts.
3. Get out of your cocoon. People tend to stick to familiar ways of doing things, and the discomfort of change often prevails over the dissatisfaction of inefficiency. Change management should be a part of the process improvement project from day one.
4. Data is everybody's business. Every business process receives, creates, modifies, uses, and transfers information. This is true for any type of business, whether material production, where information is embedded into products and parts or completely intangible businesses like finance, where information is the only object being worked on. Understanding the transformation of information at every step of the process is the key to improving its effectiveness.
5. Don’t rob Peter to pay Paul. Every organization is a complex system with many internal dependencies. No part operates in isolation. Any improvement that solves a local problem can create ripples and lead to new problems elsewhere if not considered within the context of the entire system.
6. Buy-in without automation is better than automation without buy-in. Once again, technology can only support a process, not replace it. Suppose managers and team members responsible for every function in an organization understand their roles and how they affect—and are affected by—others, and act responsibly. In that case, this will create a more positive impact than any state-of-the-art technological solution that people would struggle with.

A useful aspect of process improvement projects is that they highlight the exact places where legacy technologies have indeed become an obstacle to business growth and even provide clues as to what an organization needs to replace.