NextOrbit is a cloud platform to predict and mitigate out-of-stock.
THE PROBLEM: Fashion and Apparel brands lose 4 to 15% of revenues and be saddled with slow moving store inventory due to ineffective response to out-of-stock. Spreadsheets or rule based software may partly solve the problem. However the solution causes high inventory at stores. (Over-stock may not address out-of-stock!). You need the right products at the stores at the right time.
ACHIEVING THE BALANCE NextOrbit is a cloud platform that uses machine learning and data sciences to achieve the twin conflicting goals of out-of-stock and over-stock.
HOW WE DO IT: The first step is to obtain the signal for each product and each store/online. We account for the velocity and determine the aggregation interval most appropriate to the item and store. Once the signal for an item is obtained, we super-impose holidays, local events, weather patterns, fashion indicators, social feeds and macro-economic indicators to understand the lift or dip due to each factor.
Fashions items have new designs every season. To address new products, we obtain the signal at a higher level of the hierarchy (e.g. full sleeve dress shirts), then distribute the demand to the specific items (e.g. full sleeve pin striped light blue shirts). We create insights of what is trendy – e.g. “ in this part of the country, light blue shirts with thin black stripes are in vogue..” We determine and account for fashion fatigue when computing future demand.
OUTPUTS: Store allocations or orders, shelf-out alerts, Demand plan
BENEFITS: Customer experience and loyalty improves. There is a same-store-sales improvement of anywhere from 4 to 15%. Store inventory is optimized and inventory turns improve.
And all of this happens not just one time but in a loop of continuous improvement. Our researchers are constantly looking for new mathematical modeling methods and new predictors – factors that can determine the success of your brand.
Location: United States, Texas, Plano
Mentions in press and media 2