Turning Around a Distribution Center

When I took over a multi-million-dollar distribution center, it sat in the bottom quartile of its company's performance standings. Overtime ran high. Inventory variance didn't have an explanation. Delivery windows got missed often enough that customers noticed.

Going on three years later, we're tied for 4th out of roughly 110 distribution centers nationwide — the top 5%. Along the way the standings improved 54.5%, overtime came down 15%, and delivery-window success went up 7.3%.

People ask what system we put in. The honest answer is that the tools were mostly already there. What changed was the team using them.

1. Build the team before you touch the numbers

A struggling building usually has a culture problem before it has a metrics problem. People stop raising issues because nothing happens when they do. Supervisors manage to the month-end report because that's when they get judged.

So the first work was people work. I led a team of 30: hiring, mentoring, and training. Day to day that meant huddles and regular team meetings, one-on-ones, and getting to know every person in the building.

It also meant changing what happens after a miss. When we missed, I brought the team the hard data and treated it as our opportunity to improve, not a reason to scold or berate anyone. Once people saw a miss wouldn't turn into blame, they started raising problems on shift instead of hiding them until month-end.

If you're doing this: trust comes before tools. A team that doesn't trust each other will treat any new report as something to defend against.

2. Make the tools the team's own

The DC already had Power BI and Excel. They were used the way most buildings use them: someone runs a report, someone else gets judged by it.

I coached the team to use them as working tools. We looked at daily order monitoring together, so a disruption became a staffing decision that morning instead of a variance to explain at month-end. Supervisors learned to plan a shift in Excel the same way, with the same inputs, instead of each one keeping their own version. Streamlined workflows and that shared planning lifted warehouse efficiency roughly 10%.

The measures we steered by were simple:

• Fill: are we shipping what was ordered?
• On-time windows: are trucks hitting the customer's window?
• Labor hours against volume: are we staffing to the work, or to habit?

If you're doing this: don't start by buying or building software. Teach the team to use what they have, together, every day. A tool becomes a tool when the people on the floor reach for it on their own.

3. Plan ahead, and get the team to trust it

Overtime is usually treated as a budget problem. It's almost always a planning problem that shows up in the budget.

The change that mattered most was using the Power BI planners and Excel tools proactively, ahead of the shift, instead of explaining hours after it. The harder part wasn't the planning. It was helping the team see that planning ahead actually improves their day, and getting them to trust the process.

Overtime expense came down 15%. Inventory variance improved at the same time. That matters, because under pressure those two numbers usually move in opposite directions: cut hours and accuracy slips.

If you're doing this: plan the labor before the shift, and keep the variance numbers in view while you do it. If accuracy drops while hours drop, you've moved the cost, not removed it.

4. Fix delivery reliability route by route

Missed delivery windows looked like one problem from the top. Up close they were several smaller ones. Manifest analysis showed which routes and handoffs were failing. Driver training and tighter DOT compliance closed them. Route-delivery window success went up 7.3%.

If you're doing this: don't retrain every driver on everything. Let the data name the routes and handoffs that fail, and train to those.

The short version

1. Build the team and the culture first. Treat a miss as data, not blame.
2. Teach the team to use the tools they already have, together, daily.
3. Plan labor ahead of the shift, and help the team trust it.
4. Fix delivery reliability where the data says it's broken.

Overtime doesn't drop 15% because a report says it should. It drops because supervisors trust the numbers, the crew trusts the supervisors, and problems get raised on shift instead of discovered at month-end.

The full results are in the DC turnaround case study. If you're running a building that's stuck at the bottom of the standings, I'm glad to compare notes.