Several thousand workers who ferry groceries for Instacart are on the final day of a 72-hour strike over how the on-demand delivery app pays them. Instacart’s “shoppers” grab items off stores shelves, purchase them, and deliver the order to customers’ doors—but they have a problem with how the platform manages tipping: The default in the app is for users to leave either a $2 or a 5 percent tip for the shopper on top of the 5 percent service fee that Instacart charges every order. The workers allege that this service fee is misleading and that the 5 percent default tip option has cut into their take-home wages—which they say have shrunk because the app previously suggested a tip of 10 percent, and the company didn’t charge a service fee.
This is, once again, a story about a startup in the gig economy finding ways to shrink how much it pays the contract workers who make the service function. But it’s specifically a clash over app design: Instacart shoppers say that the user interface is causing them to earn less money.
If you make an order through Instacart, you can see what they mean. I recently downloaded the Instacart app and filled my cart with $60.74 of groceries. At checkout, Instacart suggested I tip the shopper $3.04, but it didn’t specify that the amount was 5 percent of the bill—I’d have to do the math to realize it’s far less than the customary 15 to 20 percent (or more) gratuity for food-service orders. And on top of that is another $3.04 service fee. (I didn’t complete the order.)
Instacart’s tipping policies have irked its shoppers for years. Back in 2015, the platform suggested customers tip the app’s gig workers 10 percent, according to Vanessa Bain, who started delivering for Instacart that year and is one of the current strike’s lead organizers. Since then, the platform has made all kinds of tweaks that have confused both shoppers and customers. In 2016, the app took away tipping entirely, replacing it with a 10 percent service fee, and told customers that this change would be offset by increasing the commissions shoppers made. After shopper outcry, Instacart reintroduced tipping as an option to add after the order was complete. But in 2018, the company tweaked its policy again, guaranteeing a minimum base payment of $10 per order but also counting customer tips toward that base. On Reddit and in private Facebook groups, shoppers started to commiserate. In one instance posted on the website of Working Washington, a labor organizing group, an Instacart shopper shared how they made $10.80 total on an order where the customer tipped $10. In another case, two identical orders both paid the drivers $10, despite one receiving a $2 tip and another receiving a $6 tip.
This setup—in which the bonus tacked on by customers offset the amount the company paid the workers—undermined the very idea of what a tip is, and workers loudly complained that it meant they were making much less money. In response to this outcry, by February 2019 Instacart stopped counting tips toward base pay and paid back drivers who had their tips go toward the wages Instacart was supposed to provide.
Instacart workers are unhappy with the current system of a 5 percent default tip plus a 5 percent service fee because they say they’re taking home less. In 2015, Bain told Reuters she could take home around $1,500 for a 45-hour workweek. Now, she and her husband, who also does Instacart deliveries, say that they each only bring in a few hundred dollars a week from their work on the app.
“Tiny changes in the user experience can cause huge changes in user behavior,” said Lior Strahilevitz, a law professor at the University of Chicago who recently co-authored a paper on manipulative digital design interfaces, often called “dark patterns.” When it comes to tipping, Strahilevitz said, “People like to do what’s socially acceptable and what they think people around them are doing.” So, when Instacart makes the suggested tip 5 percent instead of 10 percent and adds a 5 percent service fee, “It wouldn’t surprise me if that tweak massively changed user behavior,” Strahilevitz said. “I haven’t run an experiment on this app, but I would be really surprised if upward of three-quarters of users didn’t leave the tip as is.” Consumers might also be led to think they’re tipping 10 percent, as they were previously, Strahilevitz said, even though they’re now technically tipping half of what they used to.
A dark pattern is an element in a user interface that’s designed to manipulate users to act in a way that they don’t intend to, often by manipulating how users navigate or confusing them about what their options are. It’s unclear if Instacart’s app design is intended to reduce driver tips or confuse users about the gratuity they’re leaving, but drivers are still pointing to an effect: Their overall take-home is far lower than what it was when the default tip was 10 percent. Beyond the default tip being less, another crucial design choice might be that the app doesn’t indicate that the tip it suggests is only 5 percent; while users have the option to manually enter a higher tip, they may not realize that, by percentage, the default is paltry.
Instacart isn’t the only on-demand delivery app that has confused customers about tipping. DoorDash was the subject of a huge customer backlash earlier this year following a New York Times article that detailed how the tips customers gave to drivers was used to subsidize the drivers’ base pay, rather than being given in full to the driver as gratuity. After customer outcry, DoorDash changed its policies and started giving all tips through the app to drivers. Unlike Instacart, however, DoorDash has not said it will pay drivers back for tips that the company previously collected and counted toward driver wages.
There are other ways to design user interfaces that actually encourage people to tip more, Strahilevitz said, like showing consumers an option of tipping, 15, 18, or 20 percent on an order—similar to how Lyft and the payment service Square show customers multiple options. “One of the things we know from behavioral econ and from experimental psychology is that when you provide choices, whatever is the middle option is what most consumers gravitate toward.”
Since Instacart has no interest in treating its shoppers as full-time, salaried workers—and is seeking an exemption to a new California law that would classify many gig-economy workers as employees—its delivery people are vulnerable to things like small changes in interface design. In response to the work stoppage, Instacart has told journalists, “We take the feedback of the shopper community very seriously and remain committed to listening to and using that feedback to improve their experience.” The feedback seems pretty clear: Change your app.
