To increase forecast accuracy of top-line sales as well as Amazon purchase orders.
With such unprecedented conditions, forecasting sales in and beyond the dynamic market of 2020 has been difficult for brands. In addition, companies have had to find a better way to manage the resulting dramatic fluctuations in purchase orders from major vendors such as Amazon.
The approach & insights
Even before the pandemic, our client’s business was seeing strong YOY growth in sales. Once the pandemic conditions impacted the market, sales for both their product and the product category overall skyrocketed. However, purchase orders from Amazon had taken a dramatic swing. Fluctuations in demand at the end of the supply chain were now causing larger fluctuations upstream. Fulfilling purchase orders became a struggle of allocation across sales channels.
The Tradeswell forecasting algorithms gave the manufacturer access to real-time insights and sales forecasts from the SKU-level up to the total business. Forecast accuracy improved from 25% to 70% by using the Tradeswell automated solution.
Tradeswell was able to show the company that they needed to factor in Amazon’s current inventory levels in terms of weeks of supply when forecasting purchase orders. This allowed them to assume target weeks of supply for Amazon (around 7), and build that into their assumptions for purchase orders in the short term.
Ecommerce is a new industry after the pandemic, with considerations like competitor stock-outs and price matching continuing to drive wild fluctuations in demand. Brands must be increasingly nimble enough to understand these fluctuations down to the SKU-level while being able to forecast and plan accurately, both for the short- and the long-term.