In 2012, Musthafa PC, CEO of ID Fresh Foods, had charted out an aggressive growth plan for his small company. However, he realized that he would not be able to achieve the numbers unless he found a way to bring down the high percentage of returns and wastage of his perishable food products – nearly one-fourth of iD’s flagship product, the idli-dosa batter, was being poured down the drain at that time, causing a significant leak in margins.
The wastage of food also went against the founding team’s closely held values of sustainability and social responsibility. The main cause of the high wastage was the inability for iD fresh to predict the demand for its products, most of which were preservative-free and had a shelf life of 2-7 days.
“(Back then) our sales team would update an Excel sheet (with the sales numbers) at the end of each day. By the time we computed and analysed the data, it was more than two days old… We needed to speed up the process,” said Musthafa in an interview.
Musthafa picked Bizom, a cloud-based retail sales tool for CPG companies, to digitise their van sales operations. This automated the end to end sales execution processes and provided iD with real-time visibility into sales, inventory and returns. To accurately predict demand for every store/SKU combination Bizom deployed their “Suggested order” capability. This prescriptive analytics model brings together a vast array of datasets including secondary sales, seasonal factors, ongoing promotions, hyper-local competitive information among others to identify the factors that influence the demand and suggests the optimum order that every store should place.
Machine learning capabilities of the model means that it is always learning and improving the accuracy of its predictions. The model was deployed on Bizom platform and seamlessly provided this insight into the hands to the sales reps to execute their orders most effectively.
In a matter of weeks, ID was able to achieve 100% visibility on orders and returns, increasing the productivity of its van sales personnel and even cut costs on fuel and manpower.
Most importantly, the AI-powered Suggested Orders capability helped the team shrink wastage from 25% to 2% (the industry average is between 10-20%). An analysis of the order predictions revealed a near perfect match with actual orders (the measure shows an R2 score of 0.92).
“Even if the store owners want to buy more, we let them override the quantity only by a limited amount as we are so confident about our demand prediction analytics,” Musthafa said in the same interview.
These results helped iD write its now-legendary story of registering 10X growth in four years and are now taking their business nationally across the markets they operate in today.