McDonald’s loyalty app produced a 515-page dossier predicting a customer’s next order

Summarised by Centrist

A journalist who requested the personal information held about him by McDonald’s received a 515-page file detailing his purchases and predicting how often he would return, how much he would spend and what he would order.

The data showed how an everyday loyalty programme can become a sophisticated system for analysing and forecasting individual behaviour.

Wired journalist Reece Rogers requested the information using privacy rights available to California residents.

The file predicted he would visit McDonald’s 2.16 times over the following six weeks, spend an average of US$13.49 per visit and spend US$29.15 in total.

It ranked a large Diet Coke as the product he was most likely to order, followed by a Spicy Snack Wrap and large Grinch McShaker fries.

McDonald’s also assigned him a “customer attrition likelihood” of zero, effectively predicting he would not stop buying from the chain.

Individual purchase records are unsurprising in a loyalty programme. The more consequential disclosure was the company’s use of that history to draw inferences and make detailed predictions about future behaviour.

McDonald’s says it uses customer data to personalise services, marketing and communications, improve performance, detect fraud and provide loyalty benefits.

Its US privacy policy says information and inferences may be shared with social-media, advertising and analytics partners for targeted advertising.

Rogers subsequently asked McDonald’s to delete his information and stopped using the programme, partly to test whether the algorithm’s prediction that he would remain a customer proved correct.

Read more at Wired and McDonald’s US privacy policy

Receive our free newsletter here