Key goal
Nextail’s Initial Allocation solution helps users determine how many units of a new product should be sent to each store in each SKU before any sales history is available for that item, boosting sales to where it is more likely to happen and reducing the risk of early overstocks across the network, attending to business constraints at all times.
How do we do this?
An allocation outcome is built by conciling unconstrained demand opportunities with constrained rules. It utilizes a probabilistic hyper-local demand forecasting model to estimate demand over the initial selling period dates and a mathematical global optimization model to maximize sales opportunities, considering available purchased inventory and multiple business factors.
In this way, 2 different driver-based outcomes are provided, eligible, and editable: a demand-driven baseline proposal and a planning-driven proposal. The main difference between them lies in securing, or not, any Allocation Planning input or target. Certain business Planning inputs or targets are allowed to be optionally secured during the allocation proposal, to guarantee the retailer's decisions on top of a demand-driven baseline proposal.
Summarizing, Initial Allocation calculations run throughout the following main steps:
- Demand Forecast: machine-learning models that generate a SKU-PoS demand by estimating a Product-PoS demand by day within the whole allocation timeline period.
- Global Optimization: mathematical optimization models that maximize gross revenue by evaluating multiple factors such as selling probabilities for each single SKU-PoS-Unit, service levels (sales threshold), local selling prices, local currencies and exchange rates, warehouse inventory value, and a long list of business rules (logistics & merchandising).
- Outcome reports: allocation reports in different forms, such as detailed spreadsheet results explanation by SKU-PoS, picking order (waybills) files, and on-screen results at different aggregation and disaggregation levels.
