This Is What Happens When a Data Nerd Drives Lyft
A story about the revenue, costs, and profits of being a Lyft driver.
In April 2013, Lyft was a rather new and interesting concept. They had a couple hundred drivers in San Francisco, and were looking for more.
I was taking some time off after selling my company, and I was new to San Francisco. I figured that driving for Lyft could be a good opportunity to meet some people, explore the city, and earn some cash before jumping into my next startup.
Before going all in with Lyft, I tested it out first. My only car was a two-door coupe, but Lyft required a four-door car. I was in the market for a SUV for my wife anyways, but before I purchased one, I figured I would give Lyfting a proper test run. I borrowed my sister’s Honda Civic and gave it a trial run for two days.
As it turned out, I really enjoyed it. I personally love driving and conversing with new people, so it was a perfect marriage of the two. I went forward with the purchase of a 2013 Mazda CX-5, which is by no coincidence the most fuel efficient SUV model available in America. 18 months and 5,000 miles later, here’s a summary of my experience, revenue, and profit.
Locations
I use three services to track my driving history: Chronos, Google Location History, and MetroMile.
Chronos is an iOS app that tracks my daily movement and activity via GPS From the heat map above, you can see that most of my time is spent in SOMA and FiDi. The data was taken continuously over the last two years, so it does include locations unrelated to Lyft driving. For a more accurate data map, I turn to Google.
This is my location history only from September 20th — October 20th. In that time, 90% of my San Francisco visits have been Lyft affiliated, so the data should be more accurate than Chronos. To access your own data, log into your Google account and visit this page. If you don’t see any data, you’ll first need to give Google permission to track your location.
Finally, I use a new auto insurance provider called MetroMile. They charge me based on exactly how many miles I drive in a month. To have an accurate mileage number, they gave me a free dongle that I plugged into my car’s OBD port. It monitors my car’s location, among other nifty things, and sends the data to the cloud. Check out the video animation of my driving history here.
Revenue & Profit
Since I started, I’ve earned $11,503.56 and given 955 rides, which works out to $12.05 per ride. This number is the actual revenue that gets deposited into my bank account, after Lyft takes their 20% of total ride payments.
For the current month of October, the average ride distance was 3.35 miles. In this same month, I earned $1,300 over 41.5 hours of dedicated driver time, for an effective hourly rate of $31.32. Statistically speaking, 99% of my most recent passengers gave me a 5 star review.
When determining profit, I decided to exclude the cost of the vehicle, because after all, it is used far more for unrelated trips. The Mazda is our daily driver that we use to cart around the dog and our daughter, and we would own it even if I didn’t drive for Lyft. Using October again as a baseline, I can now calculate my profit:
Revenue:
$1,300
Expenses
Miles Driven: 491.2 Logic: 341.2 with a passenger plus an estimated 150 without a passenger.
Prorated Tire Wear: $7.86 Logic: The tires cost $800 to replace, and will last 50,000 miles, for a per mile cost of $0.016.
Prorated Maintenance: $8.10 Logic: A scheduled oil service is $75, and will last 8,000 miles. I’ve also forecasted $600 in additional service (brakes, shocks) over 80,000 miles. Combined, these two result in a per mile cost of $0.0165.
Gasoline: $55.50 Logic: My Metromile app indicates an running average of 30 MPG. I assume gas to be priced at $3.39/gallon, for a per mile cost of $0.113.
Insurance: $20.14 Logic: With MetroMile, my variable insurance cost is exactly $0.041 per mile. I also pay a base rate of $28 per month, but that is a sunk cost already accounted for on my personal budget.
Tax Deduction
(+$83.26) Logic: The IRS has an allowance for $0.565 per mile. From my 491.2 miles, I would get a reduced taxable income of $277.53. Assuming a tax bracket of 30%, my tax bill is lowered accordingly.
Bottom Line
Total Earnings: $1,291.66
Basically, all of the variable costs are offset by tax savings afforded by the IRS deduction. I have $91.60 in costs, but my tax bill at the end of the year will be $83.26 lower. This is a pretty immaterial difference. Of course, this is the intent of the allowed deduction in the first place.
To be fair, some of these are figures are calculated estimates, so the final numbers could be slightly different. However, the takeaway is that they are roughly the same as the write-off allows for. Also, I have not factored in depreciation of the vehicle at all. All else being equal, the more miles I drive, the lower its resale value will inevitably be. Given that I plan to own the car for a while, I felt that this was relatively trivial to factor in. It’s also important to note that, as the old saying goes, your mileage may vary. The profit calculation ends up being very different for a car purchased exclusively for Lyft, or one that has poor fuel economy.
Have a comment or question? Join others in the conversation by using the #Lyft955 hashtag.
Tips
All of the numbers above include any and all tips that were made through the app. (Only two passengers of my total 955 have ever offered me cash tips, most people understand that any tips are given through the app.) In October, tips accounted for 3.4% of my total earnings. Of those that tipped, the median percentage was 20% and the average was 24%. The highest tip was 71% ($5 tip for a $7 ride). Of the 102 rides I gave, 17 passengers tipped.
Pricing Fluctuations
In the heated competition between Lyft and Uber, both seem to relentlessly slash rates for passengers in order to compete on price. Some of this is absorbed by the companies themselves via a lower commission fee or even subsidized ride costs, but inevitably it can mean lower earnings for drivers. On the Uber side of things, there has been quite a bit of pushback from upset drivers who feel that they don’t have a say in pricing structure. In fact, Uber drivers are even protesting nationwide.
In my experience, the pricing game is really just semantics. Sure, Lyft may have slashed fares by 20%, but they also added PrimeTime (dynamic pricing which can skyrocket based on demand). The law of averages suggests that passengers will pay the same total amount over the period of a few months, and drivers earn the same in the long run, too. Some rides will be less than the original structure, some will be more, but in the end it evens out.
In fact, my earning stats are nearly identical today than they were 18 months ago, before the introduction of PrimeTime or the rate cuts. For my first full week way back in April 2013, I made $30.44 per hour. Not a far cry from the aforementioned $31.32 that I brought in during October 2014.
All in all, I really enjoy driving for Lyft. Although I only drive on the weekends, the opportunity cost of the time spent driving will soon be outweighed by the time that I could spend growing my newly launched startup, Whttl. I’ll be hanging up my Lyft mustache, so to speak. In the meantime, I hope that we get the chance to ride together.
For a brief period, I signed up and drove on the UberX platform, too. I intend to write a post describing the difference between Lyft and Uber, from a driver’s perspective. Follow me on Medium to be the first to read it and other posts like this one.
Lastly, if this post convinced you that driving for Lyft would be super rad, sign up to be a driver here.
If you found value in this, it would be tremendous if you scrolled down a little further and hit the “Recommend” button.
Greg Muender is the founder of Whttl, described as the “Kayak.com for the peer-to-peer economy.” Use it to compare dozens of different providers and marketplaces at once, including RelayRides, DogVacay, and HomeJoy. Drop Greg a line via greg<at>whttl/dot/com. This post was originally published on the Official Whttl Blog.
