When I joined Copperleaf, we were 4 people pioneering the idea of using AI and optimization to look at asset and portfolio management. By the time that I left, we had 150+ people across multiple continents delivering complex optimization solutions to enterprises, maintaining a healthy business and a 100% customer success rate.
That growth taught me something critical: the difference between companies that scale successfully and those that hit walls isn't about adding people or technology. It's about building transformation capabilities that actually work.
The Foundation Most Companies Skip
Most transformation projects start with technology demos and proof-of-concepts. Companies get excited about AI capabilities or optimization algorithms, but they skip the fundamental questions: How does this connect to our strategy? How does it impact our workflow?
Successful projects do something different. They start with clarity about where they want to go, then work backward to understand what needs to change.
Three Pillars That Determined Success
After watching hundreds of projects, three patterns separated the wins from the expensive failures:
Start with the End in Mind
Technology enables process change, which creates the value.
Last week I wrote about the gap between documented SOPs and operational reality. It's easy to get excited about technology capabilities without understanding real current state or where you're actually trying to go. Spending the first weeks of every engagement mapping both the strategic objectives AND the messy day-to-day reality, then identifying what process changes were required to bridge that gap really paid dividends in terms of sponsorship, change management and ultimately adoption later in the project.
If you can't articulate how your project connects to business strategy, stop.
Learn Your Data by Using It
I worked with a large utility who had just finished a multi-million dollar asset data project prior to starting to work with us. Our team was very happy with how easy and clean it was to bring that data in but unfortunately, it was missing several of the key attributes that ended up driving the decision models that the utility wanted to use.
You can't figure out data quality by analyzing it in isolation. An iterative, human-in-the-loop approach is far more effective. Start using your data immediately, learn what's broken and also what really matters by trying to solve real problems, then focus effort on fixing those specific issues. This approach is faster and more effective than trying to achieve perfect data in a vacuum.
Rollout is Everything
Strong executive sponsorship gets you started, but detailed change management gets you across the finish line. The most technically perfect solutions fail without proper rollout planning. This can happen in a 50 person organization. Projects that succeed treat change management as seriously as technical implementation.
One of the roughest projects that I ever worked on was a system destined to be used by engineers in project teams but sponsored by finance. The two teams barely spoke. We got there eventually but with a lot of rework as we looped back to pull everyone in late in the process.
Why This Matters More Than Ever
Today's AI and automation technologies amplify these challenges. Bad strategy, poor data practices, and weak rollout planning now fail faster and more expensively than ever before.
But the companies that get these fundamentals right? They're seeing transformational results because AI makes good processes exponentially better.
The integration complexity has exploded, making the iterative data approach even more critical. The companies winning today built these operational foundations first, then layered in AI to automate and scale what they'd proven worked.
What's Next
I'm now helping mid-market companies apply these lessons to build transformation capabilities that actually deliver results. The technology has evolved, but these three pillars remain the foundation of every successful implementation.
What's been the biggest gap in your transformation efforts - strategy alignment, data challenges, or rollout execution? If AI hit your workflows tomorrow, would it multiply what works—or expose what doesn't?