Arena: Market Conditions Before Sidecar
People seeking an on-demand city ride · 2012 · San Francisco, United States
Regulatory constraintsEnabling technology shift
In 2012 a San Francisco rider could hail or phone for a taxi, book a licensed black car at a premium through an app, or ride with a volunteer service that asked for donations instead of fares SC7. Carrying passengers for pay required a charter-party permit from the state regulator, and licensed taxi and limousine operators carried $750,000 to $1 million of liability insurance; the volunteer service carried none and vetted its drivers lightly SC7 SC11. Smartphones could locate a rider and a driver and take payment by card, but the law had no category for a stranger driving a private car for money SC7 SC12.
What shaped the outcome
Step 1 of 5 · 2011–13
A donation-based design put private cars into paid service
Sunil Paul had helped draft California's 2010 law on peer-to-peer carsharing, which measured profit by the year rather than by the mile SC28. In 2011 he founded Sidecar with Jahan Khanna and Adrian Fortino, whose Ann Arbor team had built a bus-arrival app SC7. Their first three ideas, a smartphone bus on a fixed route, a routed shuttle for groups of friends and shared black-car rides, were dropped as illegal or too complicated SC7. The fourth copied a volunteer car service that took donations instead of fares and fixed its weaknesses: suggested donations the rider could change, background checks to San Francisco taxi standards, company liability cover up to $1 million, card payment and ratings SC7. A test site was live by February 2012 and the public service opened in San Francisco that June; by October it counted hundreds of drivers and more than 50,000 rides SC7 SC10. The state regulator sent cease-and-desist letters in August and in November fined Sidecar and Lyft $20,000 each for running unlicensed charter services. Paul answered that the law had not foreseen the medium, and Sidecar kept operating SC10 SC11. Settlements in 2013 let the companies run while the commission wrote rules, and its September 2013 decision created a new category requiring $1 million of insurance, background checks and driver training SC12.
The design made supply cheap to add: any vetted driver with a car could join, the same lever Uber had pulled with licensed black-car drivers two years earlier, applied to a far larger pool. It was also easy to copy. Within weeks Paul saw Zimride's founders using the app, and their Lyft followed with a pink mustache on every car SC7. In September 2012 Travis Kalanick said a rival with an advantage in getting supply was a problem and that Uber had to become 'a low-cost Uber as well' SC27. Once the regulator wrote one set of rules around all three companies, the legal design no longer set Sidecar apart SC12.
Rivals Lyft launched its own peer service within weeks with a conspicuous brand, and Uber answered with a low-cost tier while it operated in 17 cities to Sidecar's one SC7 SC27 SC10.
Designed for TrustHow to win
Step 2 of 5 · 2012–15
Seven new cities in a month, served on weekend nights, never filled
Sidecar's first round, $10 million from Lightspeed and Google Ventures in October 2012, was meant for a city-by-city rollout into dense areas with constrained taxi supply SC10. In February 2013 it bought Heyride, an Austin ride-sharing startup, launched Philadelphia, Austin and Los Angeles at once, and named New York, Chicago, Boston and Washington as next SC13. The new cities ran from 5 p.m. to 3 a.m. on weekend nights until each reached critical mass. Philadelphia impounded drivers' cars in a sting, Austin ruled the service illegal and Sidecar sued the city, and free rides during South by Southwest drew more than 20,000 requests SC14. In August 2014 Paul counted thousands of weekly drivers across ten cities, recruited mostly by word of mouth and a $5 referral coupon SC9. That September Sidecar served eight metro areas, while Uber had raised $1.5 billion and served more than 205 cities and Lyft had raised $333 million for more than 60 SC15. By 2015 Sidecar offered drivers hourly guarantees if they booked three-hour blocks, completed a ride an hour and stayed inside designated areas SC23. Harry Campbell, a Los Angeles driver who wrote about the industry, said that outside San Francisco drivers 'might get a ride every hour or two' and left the app SC1.
A ride app improves for both sides only as a city fills: nearby drivers shorten waits, and steady requests keep drivers earning. Bill Gurley, an Uber board member, described that loop in 2014 SC30. Sidecar opened cities faster than it could fill them. Weekend-night hours and free festival rides seeded demand for a few hours a week, and drivers who waited an hour for a fare stopped driving, which thinned supply again SC14 SC1. Campbell counted San Francisco as the only city where Sidecar was a viable option for drivers, and the analyst Jan Dawson noted that Uber could afford to subsidize markets in their early days SC1.
Rivals Uber reached San Francisco density first and then entered cities at many times Sidecar's pace; by September 2014 it served more than 205 cities and Lyft more than 60, against Sidecar's eight SC15 SC27.
Cold StartMarketplace network effectsCustomer Referrals
Step 3 of 5 · 2014–15
Drivers set their own prices and riders chose the driver
In February 2014 Sidecar replaced company-set fares with a marketplace. Drivers set their own prices and described themselves in the app, and riders picked among nearby drivers by price, rating and distance, with the price fixed before the trip and no surge multiplier SC20 SC21. Union Square Ventures had invested about $10 million to pursue the model. Its partner Fred Wilson wrote that 'Uber is efficient and Sidecar is personal', as Amazon was to Etsy SC20 SC21. Sidecar reported a 50 percent rise in both ride requests and drivers offering rides in the markets where it tested the model SC20. In August 2014 Paul said the model returned to the company's original vision, that drivers now competed on price, and that a Wall Street Journal comparison had found Sidecar the cheapest in most cities checked SC9. He also said many of Sidecar's drivers had come over from UberX for the control it gave them SC16.
The marketplace gave drivers control over their earnings and riders a price they could see, and in the test markets it drew more of both SC20. It also asked riders to choose. Campbell said passengers found Sidecar confusing next to Uber's 'just press a button', and one reporter's search returned eleven similar offers SC1 SC20. Without surge pricing, Sidecar also gave up the lever Uber used to call drivers out when requests went unfilled, which Gurley documented from Uber's 2012 Boston test SC31 SC8. A low posted price did not bring a car to a rider in a thin market.
Rivals Uber and Lyft assigned the nearest driver automatically and set the fare centrally; both later copied Sidecar features such as driver destinations and favorite drivers SC20 SC1.
Penetration PricingCustomer Referrals
Step 4 of 5 · 2014–15
Shared rides cut the fare before the rivals offered them
Paul's original plan had been multi-passenger, multi-stop rides by everyday drivers. The 2012 launch cut it to one passenger and one destination, though every rider had to enter a destination SC7 SC2. In May 2014 Sidecar tested Shared Rides for people going the same way and counted 13,000 requests during the test. Paul said sharing let the company 'drop the price and make it into daily service' SC24. By September thousands of shared rides a week were being matched in San Francisco, and Paul put shared fares at up to half of Uber's SC22 SC26. The $15 million round that month, co-led by Union Square Ventures with Avalon Ventures and Richard Branson, was to put the feature into every Sidecar market and then add markets SC22 SC15.
Shared rides were a low-price tier like the one Uber launched in 2012, reached by pooling riders instead of using cheaper cars, and Paul saw them as the way to reach daily commuters SC24 SC27. Pooling needs even more density than single rides, because two riders must want the same route at the same moment, so the feature worked where Sidecar was already busy and could not create density elsewhere SC22 SC1. Paul himself traces UberPool and Lyft Line to the same idea; they ran on networks many times larger SC7 SC15.
Rivals UberPool and Lyft Line offered the same pooled trip on networks with more than 205 and more than 60 cities SC7 SC15.
Penetration PricingBlue Ocean Strategy
Step 5 of 5 · 2014–16
Short of capital, Sidecar moved from rides to deliveries, then closed
Sidecar raised about $35 million in all: $10 million in 2012, about $10 million from Union Square Ventures, and $15 million in September 2014 SC10 SC20 SC15. By the shutdown Lyft had raised $1.26 billion and Uber several billion SC18 SC19. In August 2014 Paul said Uber had used deceptive ride requests to recruit Sidecar's drivers, as The Verge had reported it doing to Lyft's SC16. In February 2015 Sidecar launched same-day deliveries carried in the same cars as passengers. It said a six-month Bay Area test had cut delivery cost by about 80 percent and raised driver income by 75 percent, that deliveries were already 10 percent of San Francisco volume, and that they would be half the business by the end of 2015 SC17. In August 2015 it moved most of its resources to delivery; Forbes wrote that it had judged itself too far behind Uber and Lyft to catch up SC18 SC19. Delivery put it against DoorDash, Postmates, Amazon and Uber's own courier service SC19 SC17. Sidecar ended both services on December 31, 2015, and early in 2016 General Motors bought its assets and hired its staff SC4 SC25. In December 2018 Paul sued Uber, alleging subsidies to riders and drivers meant to drive rivals out and campaigns of fraudulent ride requests; in May 2020 a federal magistrate judge let the attempted monopolization claims proceed SC5 SC3 SC29.
Paul's own account is that Sidecar tried to make ride-hailing less capital-intensive through product, and could not overcome the rivals' money to recruit drivers and riders SC7. Deliveries were meant to fill idle driver time. The August move instead took resources from rides, the only business where Sidecar had a position, and put them into a market with better-funded specialists SC18 SC19. Paul's lawsuit attributes the loss to unlawful conduct; the court has ruled only that the claims were plausible enough to proceed SC5 SC29.
Rivals Uber and Lyft kept raising: $1.5 billion and $333 million respectively by September 2014, when Sidecar closed its last $15 million round SC15.
Resource allocation