by Dharna Noor

...

The researchers then partnered with a PhD student at Massachusetts Institute of Technology, Dan Calacci, to build a bot that could analyze the screenshots Shoppers submitted, using optical character recognition. The team then used that data to assess how pay under V2 compared to the previous model.  

According to their conclusions, 41% of Shoppers are earning less under the new payment scheme. Among the pool of Shoppers who are making less, the average grocery trip is netting them 11% less than it would have under V1 pay.

The findings are consistent with Shipt’s claims that “while some orders may pay more and others pay less in the updated model, we have seen average base pay levels remain consistent overall.” In fact, the analysis found that overall, workers are making more under the new pay model (“about $0.99 more on average, per-shop”) that those gains “are not being distributed evenly”. About two fifths of Shoppers in the dataset were consistently making less under V2 than they would have under V1, suggesting that the new algorithm is “systematically under-paying certain workers.”

“Orders across the country can vary by a number of factors and our updated pay model better accounts for the effort it takes to complete and deliver orders, with factors like drive times for delivery, peak shopping windows, and the number of products in each order,” Molly Snyder, Shipt’s chief communications officer, said. “We also provide transparency to shoppers with each order—including a pay range with a guaranteed minimum pay for the shop, the delivery address, and a list of all the items being purchased—allowing them to choose whether or not they want to accept the order.”