Many years ago, I worked for a large Business Process Outsourcing (BPO) consulting firm. At the time, most outsourcing discussions focused on labor arbitrage. Move work from a high-cost region to a lower-cost region and save money. Simple enough.

What always struck me, however, was that some of the biggest savings often occurred before the process was ever outsourced. Before companies could move a process, they first had to understand it. They had to document it, explain it, and map it. In doing so, they frequently discovered inefficiencies they didn’t know existed. Duplicate activities appeared. Workarounds surfaced. Entire process steps occasionally turned out to have little reason for existing at all. Simply documenting how work was actually being performed often generated value on its own.

The same thing happens during Digital Transformation

Today, many organizations begin digital transformation by evaluating technology. Which software should we buy? Should we move to the cloud? How can we use AI? Those are all important questions. They are just not the first questions.

The first question should be much simpler: Do we actually understand how our business operates today?

Most organizations think they do until they start looking. That’s when the spreadsheets appear. That’s when the manual workarounds appear. That’s when someone discovers that a critical process exists almost entirely outside the official system of record.

I’ve seen organizations invest considerable time and money implementing new technology only to discover that they were automating processes they didn’t fully understand in the first place. Technology can improve a process, but it rarely fixes a poorly understood one. You cannot transform what you do not understand.

AI doesn’t change the fundamentals

The current AI revolution has made technology the driver behind many transformation initiatives. That’s not necessarily a bad thing. Every day we hear new stories about AI assisting workers, automating tasks, and improving decision-making. Naturally, organizations want to understand how they can take advantage of those capabilities.

However, AI still requires clean data. AI still requires connected processes. AI still requires organizations to understand what they are trying to automate in the first place.

In many ways, AI is exposing the same problems digital transformation has always exposed. Poor visibility. Disconnected systems. Incomplete data. Undocumented processes. The technology may be changing rapidly, but the foundational work remains largely the same.

Before organizations start thinking about what AI can do for them, they should probably spend some time understanding what data they have, where it resides, and whether they trust it. Knowing what to automate is often more important than deciding how to automate it.

Start there

The overarching purpose of any manufacturing company boils down to delivering the highest quality products at the lowest possible cost. Technology can help achieve that objective. AI can help achieve that objective. But before either can create meaningful value, organizations need to understand how work is actually being performed today.

Digital transformation doesn’t start with software. It doesn’t start with AI. It starts with understanding your processes, your data, and your business as it exists today. That is where meaningful transformation begins.