Six Years of Duct Tape: What We Learned Rebuilding From Scratch

We once paid developers for six years to keep a dead system alive. Vendor support had ended, and the system was customized so heavily that the vendor’s own updates wouldn’t take. So we did what a lot of companies do: kept paying developers to bend it to our will. Duct tape and prayer, at scale.

The company was $15 million in revenue when the customizations started. It grew 20x on that system before it simply couldn’t scale, report, or keep up anymore. Leadership made the call to start over with a ground-up implementation, a three-year effort.

In the first planning meetings, we heard the sentence that starts every one of these fires: we need to rebuild it to our specs. We pushed back hard. We had one shot to do this right — why spend millions recreating the exact mess that got us here?

So we did something slower and less exciting instead. We documented every process first, mapped every discipline before anyone touched the new system, then held each process up against industry best practices and asked one question: is this customization a real business requirement, or a workaround? The answer surprised everyone. Not one customization was critical. Every single one existed because the old system couldn’t do basic reporting — years and real money spent building requirements that were really just scar tissue around a broken system.

We fixed the processes and ran the new system standard. The customizations didn’t get migrated. They got deleted. That was early in our career. Twenty-five years later, we still watch companies make the same call — different system, same scar tissue. If your team is telling you ‘we can’t run standard, we’re different,’ document your processes first. You might find out you’re not that different. You’re just used to the workarounds.

Author Profile

Lee Stout