Product Portfolio Optimization

Assortment optimization works per store and shelf and decides which of the available products a given store lists. Portfolio optimization sits one level higher, typically at a manufacturer, brand owner or distributor, and decides which products exist at all. It combines sales history and margin per item after discounts and rebates with costs that an item‑level P&L does not show: forecasting effort, safety stock, production changeovers, minimum order quantities for packaging and raw materials, and warehouse locations.

Product portfolio optimization in practice

SKU profitability reports built on revenue and margin miss the safety stock that splitting demand into variants adds. Take a product line that sells 1,000 units a week with a standard deviation of 200 units. If the same demand is split evenly across four variants whose fluctuations are independent, each sells 250 units with a standard deviation of 100, so relative variability doubles from 20% to 40%. Safety stock grows with the standard deviation: at the same service level and the same total sales, four variants need protection of 4 × 100 = 400 units instead of 200. In general, n equal variants need √n times the safety stock of one aggregated item. The effect is weaker where customers switch to another variant when theirs is missing, because substitution pools part of the demand again, and strongest where variants are not interchangeable, such as pack sizes that different customers list separately.

ABC analysis by revenue ranks items by sales alone and does not show what ties an item to other business. In B2B, a low‑revenue item may be one of dozens of lines a key customer orders together; if the supplier stops carrying it, the customer has to source it elsewhere and the rest of the order is exposed. Before cutting the tail, check which customers buy the item and what share of total volume they represent, not only the item’s own revenue. Where a variant has no such ties and most of its demand moves to a neighboring variant, delisting reduces inventory and changeovers while revenue largely stays.

For retailers, the store‑level decision runs in Veritico RANGE: store clustering, product recommendations, assortment optimization within space and rules, and a KPI engine whose what‑if simulations show the effect of changes before approval, including quick listing and delisting. For manufacturers and distributors, Veritico STOCK segments the portfolio by ABC and by sales frequency into dead stock, sporadic and regular items, and handles new products through predecessor–successor links that carry over part of the sales history. See the RANGE module, inventory and replenishment and new products and seasonality.

Related terms

• Assortment Optimization
• Assortment Rationalization
• Long Tail
• ABC Analysis
• Safety Stock