The situation (2015)
Steel Dynamics is the operator most of the industry measures itself against. That winter at Fort Wayne, Indiana, storms and subzero temperatures delayed shipping for more than ten days. Production was only briefly interrupted, and the division chose to keep running. That choice is what turned weather into a crisis. Excess inventory piled up in the yards and clogged the shipping and handling processes. The division's forecast became, in its own words, unacceptable but unavoidable.
What we did
- The capability was already in the building, barely. Rob Simon had brought the profit optimization system to the division after two previous steel implementations, and the team was less than three months into an implementation that was not yet fully commissioned when the winter hit. The leadership team built the plan and owned it. The fractional executives on the engagement built the model behind it and worked alongside that team to set the conditions it ran against and to define the priorities it optimized for.
- Two days to verify the model's inputs against what the storm had actually done and to validate the results coming out of it. From there it ran nearly constantly, rebuilt as conditions moved and as the division worked its way back.
- The model decided nothing. It computed the consequences of decisions the leadership team made about what mattered: which customer commitments were immovable, which had room, and what each plant could actually do. The team supplied judgment the model could not have. The model supplied arithmetic the team could not do. Neither half produces the result alone.
The mechanism
Why it worked
The division's instinct was the right instinct for a normal week. Keep producing, catch up on shipping later. With shipping down more than ten days, producing into a blocked yard consumed the space and the handling capacity the recovery itself was going to need. Seeing that requires holding production, shipping, handling and customer commitments in view at the same time and asking what the whole system should do, rather than what each function should do.
- The answer the team reached was counterintuitive and is on the record in the paper: raise labor utilization at the constrained plant to 86 percent while holding the others at 99 percent. That cleared the shipping bottleneck and added another 20 percent in profit.
- A leadership team recovering from a shock almost always rebuilds its plan the way it built the original one. Same assumptions about what the assets will do, same ranking of what to make first, executed faster. That produces a plan to catch up, which is a smaller ambition than the plan the crisis interrupted. Beating the original forecast requires re-deciding the mix and the sequence against what is actually true now.
The sentence a chief executive should take from this page: the plan took a couple of days because the capability had been under construction for months. A division that starts building decision capability when the crisis arrives spends the crisis building it. This one spent the crisis using it, and it was not even finished.
The result
A recovery plan, created by the leadership team in a couple of days, which increased profits from the original pre-crisis forecast by more than 30 percent, in addition to achieving permanent profitability improvements.
“As hard as we might have tried, I don't think we could have found this solution using our standard practices. The new system paid for itself many times over before it was even fully commissioned.”
What they owned afterward
The permanent profitability improvements are the division's own words, and they are the part that outlived the winter. What stayed was not a plan. It was the ability to re-decide against actual conditions rather than rebuild the old plan faster. Rob Simon went on to engage Pilot Advisors at two further companies after Steel Dynamics, and today is Chief Executive Officer of JSW Steel USA, the first case study on this site.