Demand Forecasting That Cut Excess Inventory by 18%
Client
Major US specialty retailer
Industry
Retail - home goods
Services
Demand Forecasting · Data Engineering
Timeline
7 weeks
This case study presents how a leading US retail brand in home goods market partnered with Dastellar to optimise its supply chain by implementing demand forecasting solutions.
The Situation
The client was managing inventory for seasonal home goods products across multiple locations using historical sales averages and buyer intuition. Purchasing decisions were based on what sold last year, adjusted by gut feel with no statistical modelling and no systematic demand signals.
The result was a persistent dual problem: excess inventory on slow-moving SKUs tying up capital and warehouse space, while high-velocity seasonal items ran out before demand peaked. The operations team knew the problem existed but had no way to quantify where the biggest losses were or predict what was coming next quarter.
Their existing data lived in an ERP and Shopify that were functional but underutilised — transaction history, supplier lead times and seasonal patterns were all available but had never been structured for analytical use.
What We Built
We started by integrating historical sales data, supplier lead times and seasonal indicators from the client’s existing ERP and Shopify store into a clean analytical dataset. No new infrastructure was required since we worked entirely with what they had.
On that foundation we built a demand forecasting model calibrated to the client’s specific product categories and seasonal patterns. The model produces weekly and monthly SKU-level forecasts that the purchasing team can act on directly, not a dashboard they have to interpret, but specific inventory recommendations tied to predicted demand.
Alongside the model, we built an automated insights report that flags the highest-risk inventory positions each week: which SKUs are trending toward overstock, which are at risk of stockout and where the gap between current inventory and predicted demand is largest.
Results
Reduction of Excess Inventory
18
%
Reduction in Lost Sales
22
%
Supply Chain Cost Reduction
12
%
Working on a similar problem?
Start with a free iteration — we scope the problem and build the first deliverable at no cost.