Back to Case Studies
RetailSingapore

Retail Group Optimizes Inventory Across 50 Stores with Demand AI

Our client replaced static replenishment rules with demand-aware inventory forecasting. The result: fewer stockouts on top sellers and less trapped capital in slow-moving SKUs.

The Challenge

Store managers were over-ordering for safety and under-forecasting promotions. HQ teams relied on late spreadsheets, causing working capital drag and uneven in-store availability.

The Solution

We implemented demand forecasting with seasonality, promo signals, and branch-level variance. The system now recommends replenishment, inter-store transfer, and markdown timing decisions.

The Shift (Before vs. After)

Stockout rate (top 200 SKUs)

Before

14.2%

After

9.8%

Inventory holding cost

Before

$8.2M/yr

After

$7.0M/yr

Manual planning hours/week

Before

84h

After

29h

Key Results

15% reduction in total inventory holding cost
30% fewer stockout incidents on priority categories
80% of SKUs now run on semi-automated replenishment
Material improvement in cash conversion cycle

Client Perspective

"We finally have one version of inventory truth. Buyers can act on signal instead of intuition."

R. TanHead of Merchandising, Our Client

Company Profile

Client

Retail Group

Location

Singapore

Industry

Retail

Business Type

Omnichannel retail network with 50 physical stores

Implementation Timeline

10 weeks across pilot and chain rollout

Primary Impact

15% inventory cost savings

Tech Stack Used

BigQuerydbtPython forecasting servicesLooker operational dashboardsSlack alerting

Rollout Plan

  • Week 1-3: Data unification + SKU hierarchy cleanup
  • Week 4-7: Forecast model calibration by category
  • Week 8-10: Pilot to 50-store rollout with guardrails

Ready for similar results?

Let's discuss how we can apply these same strategies to your business.