May 22, 2019
The predictive elephant in the room.
More predicting. Less forgetting.
Predictive results matter.
9 min read
Why retailers will see stronger performance with predictive merchandising in their back pocket.
Your planners are awesome.
Planners and buyers are amazing people – they have the experience to make product demand forecast decisions based mostly on historical data and gut intuition. These decisions are at the beginning of the inventory cycle so they have an impact on inventory investments, allocation, marketing and ultimately product sell-through.
FIND Merchandise™ leverages the power of predictive personalization to help retailers add a third factor to their demand forecasting process – machine learning.
The advent of advanced technology like machine learning in retail has been monopolized by the marketing department – and this makes sense. Historically, marketers have been using a myriad of data points to try and predict what to sell to which customer – so when predictive personalization came to the forefront, marketers jumped on it.
Optimized buys based on historical data, gut intuition and predictive machine learning? Absolutely.
It worked well in this area because retail marketing has always been as much of a science as it is an art. Buying however, is more often looked upon as an art.
FIND understands that while predictive personalization is effective in marketing, being able to predict demand in the planning and buying stages – the beginning of the inventory cycle – would have a more substantial impact on the bottom line and on the entire organization.
With FIND Merchandise™ retailers can optimize buys based on historical data, gut intuition and machine learning to more accurately predict demand to meet and exceed their financial goals, reduce overstock and out of stock, and meet customer expectations.
We took one of our customers, a popular fashion retailer and compared their 2018 planning schedule and outcomes to what FIND Merchandise would have predicted and here are our eye-opening findings.
Our goal was to better predict how their 2018 buys could have been optimized for profit. We took 2 years of their historical sales data (2016 and 2017) and ran those numbers through FIND Merchandise™. At a high level, the sales for 2016 and 2017 fit into 2 categories – products that sold at a higher margin (above 40%) and products that sold at a lower margin (below 40%).
The vast majority of the products that were sold at the lower margins were overstock that had to be marked down in order to be sold. FIND Merchandise™ then used that historical data to predict how much inventory from each category should have been bought in 2018 … the results were shocking.
We compared their actual 2018 buys and sales with what FIND Merchandise predicted and we found that if they applied FIND Merchandise™ and moved the budget from lower margin products to higher margin products as FIND predicted, they would have increased their revenue by 3.5% ($13 million) while increasing their average margin by 10%.
Built on our award-winning recommendation engine, FIND Merchandise™ helps you buy right so you can experience an increase over last years revenue while protecting margins, improving your overall contribution and empowering you to more accurately predict what to buy, how much to buy, and where to put it. Retailers are now empowered to use the latest advances in machine learning and artificial intelligence in 5 areas of the merchandising process: