When it comes to food safety innovation, not many industry experts would consider Amazon.com a leader. The online retail giant is working diligently to change that opinion as it strives to do in fresh foods what it’s done in books and dozens of other categories.
Carletta Ooton, Amazon’s vice president, health, safety, sustainability, security and compliance (wow, that’s a mouthful), outlined the company’s approach to food safety at the 2018 Food Safety Summit, held last week in Chicago. Ms. Ooton described Amazon’s innovation as a mission critical element for the retailer, one that has the support of every level of the organization, up to and including CEO Jeff Bezos.
During her keynote, Ms. Ooton shared a few case studies to explain the company’s food safety strategy. One of the examples showed how the company uses physical process control in the deli slicing operation to reduce the risk of food borne illness and contamination. All steps — from the unwrapping of the bulk meat or cheese to the individually packaged orders being prepared for delivery — are scanned, and the data is collected to track shrink dates and alert workers to any potential issues.
A second case study involved the use of machine learning for detection and assessment. Amazon uses natural language processing to detect customer sentiment found in 16 million emails, phone calls, chats, product reviews and social media posts on a weekly basis and uses the findings to alert managers of potential contamination or other issues. For example, a series of negative product reviews allowed Amazon to detect a food safety issue and block a product 57-days before it was recalled by the manufacturer.
Ms. Ooton said Amazon views food safety as a non-competitive issue, insisting the company is working with its suppliers and others to strengthen the recall process and reduce weak links in the supply chain. She wouldn’t discuss the food safety activity at Whole Foods, the chain bought by Amazon last year, saying the systems of the two companies have not been integrated.