Case Studies / Building Products
Manufacturing · Building Products ~$240M revenue · ~600 employees · 4 plants · 2 warehouses

Building Manufacturing Group

A make-to-stock manufacturer of engineered windows, doors, and architectural millwork.

Lead service: Predictive demand & inventory modeling, delivered as a managed data service Engagement: 20 weeks

The situation

  • Demand swings hard with housing starts and the season; the plant planned by spreadsheet and gut.
  • Fast movers stocked out in peak while cash sat frozen in slow-moving finished goods.
  • Sales, production, and raw-material data never met in one place, so no forecast was trusted.
  • Long-lead components (glass, hardware) had to be committed months before demand was known.

What Grounds did

  • Modernized and reconciled sales, production, and inventory data into trusted tables.
  • Built a predictive demand forecast tied to leading indicators: housing starts, backlog, seasonality.
  • Stood up an inventory optimization model balancing service level against working capital.
  • Delivered it as a managed data service: Grounds-hosted dashboards and a live S&OP scenario simulator.

Capabilities deployed

Data modernization Predictive demand forecasting Inventory optimization S&OP scenario modeling Managed data service

Results

−41%
Stockouts on fast-moving SKUs
$9M
Working capital freed from slow movers
62→84%
Demand forecast accuracy
+12 pts
On-time-in-full delivery

What stays: A Grounds-hosted demand-and-inventory model the planning team runs every cycle.

Explore the model

Run your own numbers.

The same kind of live model we build inside the engagement. Move the inputs and watch the outcome change.

Model 1

Forecast accuracy → margin at stake

Every point of forecast error becomes lost sales or dead stock. Here's the dollar swing, and what the accuracy gain is worth.

How to use: push the forecast horizon out and raise demand volatility. The band widens and accuracy drops toward the old ~62%; the lost margin and carrying cost climb, and the green figure shows the margin Grounds' accuracy gain recovers at this setting.
4 wk
25
Accuracy 84% at a 4-wk horizon, up from ~62% before Grounds · error costs ~$3.68M/yr on ~$240.0M of demand.
$2.28M
Lost gross margin from stockouts / yr
$1.40M
Excess-inventory carrying cost / yr
$6.50M
Margin recovered vs pre-Grounds 62% / yr

Illustrative example, not a specific client engagement.

Model 2

SKU inventory policy → working capital

Set service by SKU class, not across the board, and watch inventory, COGS turns, and the cash it frees.

How to use: the old way held 98% service on every SKU with long lead times. Switch to ABC-differentiated service, set the A-item target and replenishment lead, and see inventory (cycle + safety stock) by class, and the working capital freed against that pre-Grounds baseline.
96%
3 wk
ABC holds 96% on A-items and steps down the tail: about 8.3× inventory turns on ~$168.0M of COGS, up from ~5.6× under the flat policy.
$20.3M
Inventory on hand · 8.3× turns
$4.47M
Annual carrying cost
$9.47M
Working capital freed vs old policy

Illustrative example, not a specific client engagement.

More in Manufacturing

Packaging — Manufacturing Packaging Company

Recognize your business here?

Tell us your industry and the decision you're sitting on. We'll show you the version of this model built around your numbers.

Start with your industry →