Five Ways to Improve Uber or Lyft
As background, I’ve never worked at one of these companies, so my opinions come purely from the view of a consumer. There are lots of smart people at both companies working on a slew of things, so mostly I’m just curious if any of these things are on the roadmap or have been ruled out. And final caveat: regulatory issues play a huge role in decisions they have to make, and here I’m just focused on the consumer product side of things. That said, here are five things I’d do:
Problem #1: Non-Optimal Driver Utilization
1. Reduce Driver Downtime
At the heart of it, these companies are operations and logistics companies — a key component of their success is optimizing for speed of pickup and speed of route. The speed of route is usually outsourced to Waze or Google Maps, which seem to do a fine job, but speed of pickup seems broken.
From what I can tell, here’s a typical driver cycle after pick-up: 1) drive passenger to destination; 2) drop-off, mark as dropped off in app; 3) drive around, wait for ride request; 4) opt-in to select that person for pickup; 5) go pick them up, and then back to step 1. You want drivers to have a passenger in their car for the highest percentage of their time on the road, and steps 3 and 4 hinder that ability.
Why not have a driver opt-in to a shift, and while they’re working, the app just tells them immediately what their next pickup is after they’ve done a drop-off? Passengers aren’t left waiting as long for a match in-app—you could match them with a driver on his or her way to a nearby destination, and bake in the time for the dropoff—and you avoid the driver having to operate their phone to pick out a person while driving around aimlessly.
note: I originally wrote the above on 1/23, and since then, Lyft has done some interesting things with triple matches in Lyft Line, which addresses this pretty well for carpools. Also, I suspect the issue here unfortunately is in navigating the legality of drivers not being employees, but rather opt-in contractors, and having them opt-in for a shift versus for each ride might affect that categorization. Ironically, it would be better for all parties involved (drivers, riders, parent company — even other cars on the road, for safety) to have less driver downtime.
2. Allow scheduled pick-ups
Similar to above, these companies are rooted in operations and logistics, and it’s much tougher to be optimally efficient when you’re getting all your information in real-time — i.e. every request coming into the system is “Pick me up ASAP.”
While the initial magic of these apps was the ability to summon a car in minutes, at this point, anyone who’s used it once has already had that experience. For a consumer, there are a number of times when I’d like to book in advance — heading to a meeting, the airport, pretty much any time when I’ve got hard arrival time I need to meet.
By doing this, you could not only likely be more operationally efficient, but you could also do interesting consumer-facing things. Offer an opt-in to sync with iOS calendar and let people choose at the beginning of the day or week which meetings they’d like to advance book trips for. Sync with TripIt to figure out when they might need airport rides. With the traffic data, they could even propose pick up times based on expected travel time, with a user-adjustable cushion.
note: This feature may have to come with a slight price bump, since if you’re heading to an important meeting, you can’t have a driver showing up 5 minutes late. By the same token, you don’t want drivers showing up too early and then having additional downtime, and a Supershuttle-style window is a bad experience. So perhaps you’d pay an additional couple bucks for a scheduled ride, essentially paying for the additional driver downtime.
Problem #2: Stickiness
3. Create Loyalty Programs
This one is the most befuddling to me. In fairness, Uber did launch something like this, Elite VIP Status, that I didn’t hear much about and it was unclear that the rewards were actually things people wanted. Uber and Lyft give relatively large discounts for bringing new people on board (~$10-20 on both sides), but in a city like SF where there are multiple options, it seems strange not to offer something to make switching costs higher.
A frequent rider program that either saves you money (a discount on surge? a free airport ride if you hit 20 rides in one month?) or pushes you to take an extra ride or two a month to get status, much like an airline loyalty program, could be huge in increasing stickiness. The reward doesn’t even necessarily need to be monetary, but it seems like there could be interesting levers to play with in doing more to reward riders who are frequent users, and even throwing riders’ ratings into the mix and considering that as in input for the loyalty program.
It’s possible that most people just use one app, and the rider loyalty program would just serve to cut in to profits versus engaging riders who switch between the multiple services. If stickiness isn’t a problem for the larger companies (ie very few people are switching between multiple services or abandon a given service), then this probably doesn’t make as much sense.
4. Customize the Experience
Uber recently did the Spotify partnership, which was neat, but requires a paid Spotify account, which limited it for most. Other than that, there’s not much that’s custom about a ride.
Quick note before going on: yes, it does feel ridiculous to talk about customizing a service where you literally push a button and a car shows up. Putting aside the feeling of playing into a “everything’s amazing and nobody’s happy” narrative, there are a couple levers that riders could set that would probably make for a better ride for both riders and drivers.
Some people—riders and drivers alike—view rides as a great way to connect with someone else, while others who relish the peace and quiet of their ride would rather take the time to themselves. For awhile, this differentiation was a marked distinction between Uber and Lyft, and that seems to be less so now, but even so, allowing a rider to give some indication of their personality type / preference for their preferred level of interaction could go a long way in ensuring that riders and drivers alike are more likely to have a positive experience on a given ride.
Level of interaction is probably the easiest customization here with the biggest impact, but even things like pulling your favorite music from Facebook and playing a Pandora station based on it would be pretty easy and a pretty cool experience to hop into a car that’s playing one of your favorite songs when you get in.
Regardless of type of customization, the real point here is that it shouldn’t necessarily be a one size fits all experience.
Problem #3: Leverage Their Data to Create New Experiences
5. Own Local Event Recommendations
This one’s admittedly less of a sure thing to me than the previous four, but hear me out. On a given Friday night in San Francisco, there are thousands of people going out all over the city, and before they decide to go out, the first question they have amongst their group of friends is always going to be, “Where are we going?”
At the risk of oversimplifying things, current local recommendations are divided between place recommendations (Yelp, foursquare, 7x7, Thrillist) and event recommendations (Upout, Nudge, YPlan, Sosh, funcheapSF, SF Station, Do415). Surprisingly, none of these has won, and my less-techie friends are only vaguely familiar with event recommendation apps/sites.
Given that most people don’t have a great mobile-first solution to answer the question of, “What should we do tonight?”, the door is wide open for a Lyft or Uber to more directly partner with either recommendation services or directly with events and bars/restaurants to offer recommendations on what to do that night. Granted, this is outside their current core competency, but hear my out.
I’m imagining you open Lyft or Uber on Friday at 8:00p, and a header says “Looking for something to do tonight?”, and when you select it there’s a curated list of a couple things, along with discounts offered by those institutions on your ride there. They could price it dynamically to bring more people there earlier in the night versus later, offer extra perks for bringing friends, or pair your ride with a free drink upon arrival.
Given Uber and Lyft are transporting a sizable portion of the 22–35 crowd with disposable income across San Francisco on weekend nights, there’s a massive opportunity to play a role in helping direct them towards places and events that are eager to capture their attention — and if done right, could be a major value-add for riders. Maybe event recommendations isn’t exactly it, but there are undoubtedly a couple awesome things they could do with the data they have and role they play.
Uber and Lyft have both grown incredibly quickly to become $1B+ companies, and I’ve been impressed with how quickly they’ve both been able to innovate. Their plates are full with a number of things they’re experimenting with (providing movers on demand, doubling down on logistics surrounding transporting goods, partnering with city governments, edging up towards Amazon/Google/local goods delivery, etc.) that could very well be substantial parts of their business in years to come.
Even so, I think there are some awesome things they could do to optimize driver utilization, increase customer stickiness, and leverage their data to create cool new experiences for customers. I don’t know if any of the above five are them, but as a rider, I would love to see any of these things come to fruition in the future
When he’s not writing about himself in the third person or other companies, Charlie Kubal is a co-founder of Waldo, an iOS messaging app that helps you stay in rhythm with your most important people by showing what they’re up to in the moment, whether they’re across the street or across the country, simply by carrying their phones with them.
If you enjoyed reading this, please click “Recommend” below.
This will help to share the story with others.