How AI Helps Retailers Predict The Success Of New Product Lines

Xineoh brings efficient Deep Neural Net (NN) – based consumer behaviour prediction AI to SA cost-effectively, quickly and with minimum complexity.

0
Smart retail
Smart retail: Zapp2Photo / Shutterstock.com

All retailers have models for predicting market response to new products, but with cutting-edge new fashion items and revolutionary products that have no historical sales data, it can be difficult to gauge the market success the items might enjoy.

Artificial Intelligence (AI) has the answers, says Vian Chinner, CEO of local start-up Xineoh.

AI tools – the evolution of early recommendation engines – are already delivering on their potential to accurately predict customer behaviour, improve retail sales and enhance customer experience.

Born-in-SA Xineoh, whose team of data scientists has gained global acclaim for their pioneering predictive algorithms, is already helping South African organisations improve their ability to predict customer behaviour.

In fact, Chinner is so confident of the solution’s ability to deliver measurable improvements that he offers new customers a money-back guarantee of satisfaction.

“None of our customers have asked for their money back, but a couple of them have offered to buy the company after using our solution,” says Chinner.

AI-enabled predictive analytics allows retailers to look beyond historical sales patterns to understand the characteristics of each product, along with customer affinities and spending propensity. This allows retailers to identify optimal customer segmentation; predict and prevent churn; predict what consumers will buy, and the right price to charge each user to maximise revenue.

In the case of products that are new on the market, AI helps by extracting abstractions (common underlying patterns) from data.

“Our algorithm pulls the item of clothing into components – some too subtle for humans to identify – to predict which customers will likely buy them, where sales will be best and how various stores should be stocked.”

Some of our clients in brick and mortar retail are achieving 50% improvements in forecasting accuracy over traditional models like ARMINA (statistical time series models) using our new solution, Chinner says.

Xineoh brings efficient Deep Neural Net (NN) – based consumer behaviour prediction AI to SA cost-effectively, quickly and with minimum complexity.

Leave a Reply

This site uses Akismet to reduce spam. Learn how your comment data is processed.