Inside Lyft’s Customer-Centric Growth Into a Rideshare Unicorn 🦄
Ryan George gets numbers. Before attending business school at Yale’s School of Management, he worked as a data scientist at Coursera and Lyft — two of Silicon Valley’s largest companies.
Last week, we met over a cup of tea at Blue State Coffee on campus, walled in from the biting New Haven winter cold. We talked about Lyft’s early days, subsequent growth, and how crucial their story became in later years. For a young CEO building his own shared economy startup — Homecooked: Social Dining — Ryan’s insights were so invaluable that I’ve published snippets from our conversation here.
Me: Thanks for taking the time, Ryan. One of the first things I was wondering was how Lyft thought about its first days and building the rideshare experience.
Lyft was extremely selective with their drivers early on. We started with people who were motivated to drive for reasons other than money. From a pure numbers standpoint, it didn’t make a lot of sense. But finding those people got a culture rolling.
Lyft was focused on creating peak experience at every part of the business. That started with the employee experience. We had a Willy Wonka room. We had a bunch of paintings lining a hallway, and you’d push the giant portrait of Willy Wonka and it was a door. Inside was a room with really nice carpets, high bookshelves. Why? It just made work a really cool place to hang out. But it also created this culture that made its way all the way down to the experience.
We’d go wash drivers’ cars. We’d bring in the drivers to meet each other and throw these events just for them. And that care would then filter down to the passengers, where all of a sudden you had drivers who would go out of their way to do something really exceptional for their passengers.
Me: So did that focus on experience translate to greater customer retention?
Definitely on the rider side. Before I ever actually worked at Lyft, I would always choose Lyft over Uber while I was living in San Francisco. And that’s because this idea of going out of their way to do something exceptional came all the way down from the employee experience to the driver experience and finally to the passengers. For one out of every 10 rides I took, something would happen that made me go, “Wow, I can’t believe that they just did that for me.”
“I was always willing to pay extra to ride Lyft because of that idea of something special.”
Measuring retention for shared economy businesses is hard. If you just look at week-to-week retention, it looks like passengers are churning all the time. Even if you add in the resurrected users (those that came back after a usage break), the retention numbers still looked a little problematic.
But if you look on as longer timescale, you see that the product is incredibly sticky. It’s called ‘Cohort Analysis’, where you look at a vintage of users that signed up in a specific timeframe and track them over months and years. And the surprising thing was that a really big proportion of Lyft users just stay on the platform forever — we’ve won them for life.
Me: So in the early days of a shared economy startup like Homecooked, what should we focus on measuring?
LTV — Lifetime Value. Here’s a basic formula:
Any time you can increase any of those, you’ve won.
Driver-side retention is really hard. It turns out that driving for Lyft is very specific to certain stages of peoples’ lives — they might be between jobs, just out of school, or just looking for something new.
Me: So I guess that’s why in our case culinary schools would be a really big win because that would create a steady pipeline of home cooks.
Exactly. We’d retain drivers for an average of 4 solid months and then there would be a steep drop-off. So we’d always be replenishing the supply side of the marketplace while the demand side would keep coming back.
Me: How did Lyft think about that marketplace balance as they built the driver and rider sides?
Lyft didn’t think about market balance early on — which now strikes me as incredible. They’d market too much to passengers in one area, but then they wouldn’t have any drivers. Then, they’d market for drivers too much, and then we’d have a bunch of people driving around without having any passengers for their rides.
“You need to find the right balance in order to keep both sides of the market happy.”
Me: Right now, we’re at a point where numbers are nice to validate, but not yet at a scalable marketing phase. What’s the most important thing for us, or any shared economy startup, to do now?
You need to talk to people all the time. Hear peoples’ stories on why this has made a difference, especially chefs. Get really good at telling their stories. At the crux of Lyft, the driver side was most important because you’re trying to convince people to spend all day on this platform.
Here’s the most interesting way I’ve heard anyone think about it. I have a friend who started Hitch, which Lyft acquired and turned into Lyft Line. Even more than Lyft or Uber, Hitch was the origin point of true ridesharing. That guy spent an immense amount of time with the drivers. He understood more about them and who they were than maybe anyone else that he knew! When you start small, you really have to treasure the people that believe in you and treat them like a million dollars.
Lyft created incredibly detailed personas for both their drivers and riders. Plaftorm interaction characteristics, specific marketing channels that work best, and stories around those people.
Me: After initial market traction, how did Lyft think about early growth?
Referrals on both sides are key. That’s how you drive growth. Word-of-mouth will do more for your business than anything else in this. And a referral program is just engineered word-of-mouth. If you can get people to rave about the experience, you’ve won.
But the risk of referral programs is that done wrong, they can be EXTREMELY unprofitable. A simple example can illustrate this. Say we give a user their first ride free up to $20 — sounds like a very reasonable intro offer. Lyft gives 75% of the fare to drivers, so it just cost us $15. With customer support, overhead, etc., maybe we make $2 (purely fictional — and I have no idea where Lyft is on this now) on each future ride — so we need the passenger to take 8 more rides just to break even. That cuts into profitability quite a bit even if it brings in a lot of users. Promos are often what generate the latter half of the whale curve (cf. here) through users that remove value from the platform.
Me: How does Lyft grapple with Uber in the competitive market?
I was measuring competitive market share for Lyft. During one quarterly meeting, they announced their big strategic plan for the quarter. Instead of some brilliant tactical plan for undermining Uber in key markets, it was a video of this grandmother driving for Lyft and getting stuck in the snow, and the passengers saw that she wasn’t moving on the app. They called to check in on her, and then went to help her get her car out of the snow.
I thought at the time:
“Man, that’s their strategic plan? They really have to pay more attention to the numbers than just this wild story!”
But then Trump got elected. #DeleteUber happened. And then all of a sudden, the amount that Lyft had invested in being a good — morally good — company did more for their market share than anything that I had ever done with the numbers. Lyft continued that by donating a million dollars to the ACLU (American Civil Liberties Union) but made way more back over the next months. I know that we started this conversation about numbers, but the simple fact remains:
“The story mattered more than the numbers.”
We met as part of my work on Homecooked: a social dining app that organizes small communal meals at the homes of local hosts. Although we met to discuss shared economy startup growth, I hope that this conversation will hold valuable insights for any reader. Thank you.