Fifteen years ago, an industrial organization might buy a software system and expect it to last for decades. That timeline impacted every part of the process: requirements might take a year or more to gather. Building the software often took several years. Training could take another 18 months.

All this time, however, was a good investment, because the software system had longevity. It improved operations in a measurable way and would continue to do so for years. It made sense to be 99% sure of what you wanted before moving forward.

That world is gone. Imagine if you were just now rolling out software you started scoping in 2021, before the launch of ChatGPT. The rate of change for technology today means it’s no longer realistic to expect a software system to last decades, which means it’s no longer reasonable to spend years developing one.

Instead, industrial leaders can apply the continuous improvement mindset they use to guide so much of their operations. In this piece, I’ll explain what that looks like and what the benefits are for industrial organizations that do it.

Why aren’t industrial organizations applying the continuous improvement mindset to software today?

First of all, industrial organizations are great at continuous improvement in some domains: training, safety, machinery upgrades. The problem is not a lack of continuous improvement, it’s that most industrial organizations have not yet extended that mindset to their software infrastructure.

Why not? I suspect there are two main reasons. First, software is much less visible than what's on the factory floor. You can see days without accidents rise when the team gets new welding glasses. You can see the competitive value of a new piece of towing equipment that lets you get things to customers two days earlier.

It’s harder to see what software delivers. Maybe the software was what made it possible to recognize that the old welding glasses were the highest-priority PPE upgrade. Maybe it was software that identified reduced tow time as a customer-facing metric with a big ROI. But it’s possible that only a handful of people know that or see the role the software played.

The other thing that tends to happen is that, in industrial settings, many people interact with software in infrequent bursts. When that’s the case, the software itself feels newer and less familiar. There's more of a struggle to use it correctly.

For a worker, the experience may be that, just as they’re getting comfortable with one piece of software, they’re being asked to move to a new piece.

That means the positive business impacts remain invisible and the negative impacts (like requiring non-productive training time) are what people experience. This contributes to a mindset of software as a cost center rather than a driver of business value.

What does it look like to have a continuous improvement mindset for software?

People tend to manage risk and change in what they see as more immovable structures. In industrial settings, that’s often how software is viewed: durable and largely fixed. The instinct is to manage risk rather than look for opportunity.

To move toward a continuous-improvement mindset, you have to think about software differently. Because of the tools available today, it often makes sense to build not with the expectation that a system will last for decades but for the needs your organization has today plus a little beyond.

With that framework as your starting point, you can speed up every part of the software-building process, from requirements gathering to QA. And of course, the actual code generation portion can move faster thanks to generative AI tools.

This is in some ways a business necessity for organizations operating in today’s reality. The technology available to us is evolving so fast that it often doesn’t make sense to build huge programs designed for the long term when they’ll be outdated the day they launch.

The second mindset shift is to actively tie your organization’s software to the business benefits it delivers. Does the warehouse monitoring app prevent out-of-range humidity and temperature, thus preventing the deterioration of component parts that would lead to product failures down the line?

Does the streamlined billing system reduce manual errors for the admin team? Does it get client approval for scheduled maintenance faster? All of these things have a material impact on organizational finances. Highlighting them in company updates and earnings reports helps reinforce that your software systems are drivers of institutional value.

Why shift to continuous improvement mode for software

Think of your safety program: it doesn’t make sense to wait five years and update all your safety procedures at once. FIrst, you’ll probably have a lot of preventable accidents. Second, it’s much harder for everyone to learn and retain 50 new protocols in one fell swoop. More than that, though, your team can get a lot more done if they’re healthy and safe.

Software is similar: when you evolve iteratively rather than in dramatic bursts, you make it easier for your team to learn new ways of doing things because they’re learning smaller changes more frequently. You make it possible to take advantage of opportunities that compound on each other.

Maybe you streamline billing codes and client approvals and you shave a few days off of every order cycle. Over the course of a year, that allows you to serve more customers, which increases revenue by, say, nine percent. What does that additional revenue allow you to do?

Embracing continuous improvement for software requires a mindset shift

Given the realities of how fast technology evolves today, it’s no longer realistic to wait until we’re 99% sure what software we need to build before we build it. Today, it’s possible to build something quickly, determine whether it meets your needs, and iterate if it doesn’t. 

When industrial organizations apply to software the same mindset they bring to safety practices, training programs, and machinery, they open up a new area of growth and position themselves to adapt faster and more successfully to the technology landscape as it exists today.