As consumers, we are constantly making choices about where to buy stuff we need, be it Walmart versus Target or McDonalds versus Burger King.
Increasingly, companies are creating membership programs to invoke customer loyalty. After all, a sticky consumer is a consumer for life. Just look at Costco, many consumers are more than willing to pay for annual memberships in order to gain access to the bulk quantities, low prices, and large selection of supermarket and department store products that differentiate Costco from its “big box” competitors.
And it is a profitable model—in 2018, for instance, Costco’s net earnings topped $3 billion.
When I was the Chief Economist at Lyft, we came to understand that the best way to maneuver away from cutthroat price competition with a formidable foe like Uber is to focus on something other than price. Membership programs are one option.
In Lyft’s case, the membership program would make one set of prices available to everyone and offer riders the option to buy a membership and gain access to discounted prices.
As a general economic rule, consumers are willing to pay for a membership as long as the benefits they gain from having the membership are worth more than the cost of it.
In this sense, there are two types of consumers who consistently buy into membership programs.
For the first type—let’s call them JoGoods—the better deals incentivize them to purchase even more products (or more rides, in Lyft’s case). Psychologically, the more they take advantage of the discount, the more the initial membership fee feels worthwhile, even if they are actually spending more than they would have otherwise.
This is the economic sweet spot for membership programs—consumers get a good deal, and enjoy the product or service even more. A true win-win all the way to the bottom line.
However, there is also a second type of Lyft customer, whom I’ll call NoGoods. They buy the membership because it is a good deal, but unlike JoGoods, they don’t increase their number of trips.
In their case, the membership is valuable because they ride a lot, and the discount applies to all of the purchases they would have made anyway. This is the unsweet spot for Lyft: people who are taking the same number of trips but paying less for each of them, and the membership fee Lyft collects from the NoGoods doesn’t make up for it.
The challenge for any company that wants to provide a viable membership program is the same: there must be an attractive ratio of JoGoods to NoGoods, otherwise you will lose money.
How can you determine this ratio? Do a scientific experiment!
That is exactly what we did at Lyft. We gave roughly 1.2 million people the opportunity to buy one of various membership deals.
We then watched Lyft riders who took up one of these offers crisscross their cities, leaving behind a trail of data. Untangling the patterns behind each group’s behaviors would be the key to determining if the membership model made sense for Lyft.
What did we find? The NoGoods were nearly three times more prevalent than the JoGoods, meaning that the vast majority of core customers who bought memberships didn’t increase their number of rides.
Rather, they were taking the same number of trips but at a discounted fare (a good deal from their point of view, but not so great for Lyft). The data suggested that at a three-to-one ratio of JoGoods to NoGoods, the more we expanded the program, the more money we’d lose on the NoGoods relative to what we’d earn from the JoGoods. This would be unscalable.
Going back to basic economics helps us understand when membership programs make sense. In the Lyft case, the data told us that we needed to change the benefit-cost calculus, either by changing the up-front cost or by making the membership benefits more attractive.
Thinking at a more basic level, one might wonder whether a membership program makes sense at all for Lyft. Via an economic lens, Lyft is very different from companies like Costco or Netflix, who have viable membership programs.
The data seem to suggest that a subscription model could scale only if it gave access to goods or services that weren’t readily available and accessible without a subscription.
That was the secret sauce of successes like Costco and Sam’s Club, where shoppers need to present a membership card to get in the door, and it was at odds with the ride-share market, where an Uber or Lyft arrives in seconds at the touch of a screen and riders can easily access other low-cost options like a taxi or train.
In a nutshell, as I contemplate buying a pound of chicken breast at Sam’s Club, I cannot tap an app and have Costco put a chicken breast in my basket.
Moreover, because of the economics for food products and department store goods, prices and products tend to be the same everywhere all the time, which is how Sam’s Club ensures that the selection and savings it offers are not readily available at other retailers.
Ride-sharing, meanwhile, is a dynamic product operating in micro-markets where price, demand, supply (of passenger-ready cars), and arrival times fluctuate constantly.
If one company’s supply of drivers thins out so its fares surge or wait times balloon, it’s very easy for people to switch to a competitor. In fact, many people “dual-app”—they have both the Uber and Lyft apps on their phones, and swiftly check the wait times and price on both before making the choice.
Switching is nearly effortless; you just tap a different app on your phone for an estimate, then request a ride. Tap—tap—done. Sam’s Club and Costco’s models don’t function this way, and neither do subscription (read: membership) services like Netflix, where viewers generally don’t impulsively cancel their subscription and switch to another streaming service on the spot—you have to go through the process of setting up an account and entering credit card information.
Plus, with streaming, people often pay for more than one platform simultaneously in an effort to receive a richer set of choices.
Likewise, just because you’re a member of Costco doesn’t mean you don’t also shop elsewhere. You might buy all of your household items at Costco but buy your groceries at stores with other selections of products or brands.
With ride-sharing, there is no such differentiation. In many cases, the car that pulls up has both Lyft and Uber stickers on its windshield. The drivers themselves are dual-apping!
And while even products as similar as Coke and Pepsi have different branding, the fact that Lyft and Uber drivers largely don’t makes it that much harder—but also that much more important—to inspire loyalty in customers.
As the usual saying goes, caveat emptor….which protects the buyer….but in this case we need to think about protecting the other side of the market. Caveat venditor.
Sun Prairie High School alum John A. List (@Econ_4_Everyone) is the Kenneth C. Griffin Distinguished Service Professor of Economics at the University of Chicago, Chief Economist at Walmart (formerly in the same role at Uber then Lyft). This essay is adapted from his book, “The Voltage Effect: How to Make Good Ideas Great and Great Ideas Scale,” published Feb. 1 by Currency.