How Machine Learning with TensorFlow Enabled Mobile Proof-Of-Purchase at Coca-Cola

Coca-Cola leveraged TensorFlow and convolutional neural networks (CNNs) to develop an innovative mobile proof-of-purchase system by recognizing product codes through images, significantly enhancing their loyalty program. Initially, the company faced challenges with off-the-shelf optical character recognition (OCR) tools due to low-resolution dot-matrix fonts used in product packaging, which prompted the need for a custom-built solution. By training CNNs with a combination of simulated and real-world product code images, Coca-Cola achieved a 96% recognition success rate, facilitating a frictionless consumer experience and substantial cost savings by avoiding printer updates for higher-fidelity fonts. This AI-driven solution has become integral to Coca-Cola's promotions, registering over 180,000 scanned codes since its launch in 2017. The successful implementation of this technology exemplifies Coca-Cola's shift to a mobile-first marketing strategy and opens the door for further AI applications in various business areas.