Modern MBA

Case study — Technology · 11 min read · 5 questions

Why the scooter startups all went bankrupt

The thesis

The pitch was that scooters would disrupt walking. That is not hindsight — it is what the venture capitalists wrote down. The playbook came from Uber and Airbnb: price below cost so consumers learn you as the cheap option, deploy without asking, settle with the city afterwards. Bird dumped scooters onto Santa Monica in late 2017 and became the fastest company in history to a $1 billion valuation. Uber bought JUMP for $200 million on $17 million of revenue. Ford bought Spin for $100 million.

Underneath it, the asset never worked. Bird claimed six rides per scooter per day; its filings show 2.8, then 2.5, 1.3, 1.6 and 1.3. At a 12-minute average trip that is a vehicle earning for twelve minutes and idle for the rest, while its battery drains and someone collects it overnight. The average scooter grossed $2,521 in 2018 and less every year after, against hardware at $550 to $2,479 and maintenance near $1,000 a year. It lost money on every vehicle, in every year, at every fleet size.

So Bird moved the losses rather than fixing them. Gig chargers, then Fleet Managers — handed 100 scooters free, paid 80% of ride revenue, liable for availability. That is franchising with the franchisee holding the depreciation, and it still only got gross margin to 50%. To break even at actual usage a rider would pay $8.00 a trip against the $4.97 they pay, which is an Uber. The founders and the VCs cashed out at IPO and now work at AI startups. Bird went bankrupt in 2023, and cities function exactly as before.

How do you think about this? 5 strategy questions this case raises and does not answer.
Read the comments, or add yours →

The statistics

1.3Rides per Bird scooter per day, against the six it claimed
$8.00What a ride would have to cost to break even. It costs $4.97
3–9%Gross margin on selling the scooters themselves

By the numbers — swipe or use arrows

01The playbook they copied, in dollarsOperating income in millions. Uber and Airbnb spent six years teaching consumers they were the cheap option, then raised prices as public companies.
The playbook they copied, in dollars — Why the scooter startups all went bankrupt−$10,000M−$7,500M−$5,000M−$2,500M$0M$2,500M−$3,023M−$133M2016−$4,080M−$81M2017−$3,033M$19M2018−$8,596M−$501M2019−$4,863M−$3,590M2020−$3,834M$429M2021−$1,832M$1,802M2022UBERAIRBNBModern MBA
View data
UberAirbnb
2016−$3,023M−$133M
2017−$4,080M−$81M
2018−$3,033M$19M
2019−$8,596M−$501M
2020−$4,863M−$3,590M
2021−$3,834M$429M
2022−$1,832M$1,802M
02And then the prices went upAverage cost of an Airbnb reservation and an Uber trip. Cheap was the acquisition strategy, not the product.
And then the prices went up — Why the scooter startups all went bankrupt$0$50$100$150$200$113$92017$118$92018$116$102019$124$122020$156$142021$160$152022$164$152023AIRBNB RESERVATIONUBER TRIPModern MBA
View data
Airbnb reservationUber trip
2017$113$9
2018$118$9
2019$116$10
2020$124$12
2021$156$14
2022$160$15
2023$164$15
03Half a billion dollars for no moat at allFunds raised by each startup in millions of dollars, 2017 to 2019. The scooters were the same generic Chinese model.
Half a billion dollars for no moat at all — Why the scooter startups all went bankrupt$0M$200M$400M$600M$550MBird 2018Bird2018$405MLime 2018Lime2018$310MLime 2019Lime2019$275MBird 2019Bird2019$131MSkip 2018Skip2018$100MLime 2017Lime2017$15MBird 2017Bird2017$8MSpin 2017Spin2017Modern MBA
View data
Value
Bird 2018$550M
Lime 2018$405M
Lime 2019$310M
Bird 2019$275M
Skip 2018$131M
Lime 2017$100M
Bird 2017$15M
Spin 2017$8M
04Fastest company in history to a billion, and backBird valuation in billions of dollars, from launch to year five.
Fastest company in history to a billion, and back — Why the scooter startups all went bankrupt$0.0B$1.0B$2.0B$3.0B$0.0B2017Launch$2.0B2018Year 1$2.5B2019Year 2$2.3B2021IPO$0.1B2022Year 5Modern MBA
View data
Value
2017 Launch$0.0B
2018 Year 1$2.0B
2019 Year 2$2.5B
2021 IPO$2.3B
2022 Year 5$0.1B
05250 scooters became 98,800Number of Bird scooters deployed worldwide, across roughly 400 cities.
250 scooters became 98,800 — Why the scooter startups all went bankrupt020,00040,00060,00080,000100,000120,0002502017Launch23,0002018Year 144,0002019Year 238,0002020Year 368,6002021IPO98,8002022Year 5Modern MBA
View data
Value
2017 Launch250
2018 Year 123,000
2019 Year 244,000
2020 Year 338,000
2021 IPO68,600
2022 Year 598,800
06The more they deployed, the less each one was usedAverage number of rides per deployed vehicle per day. Bird publicly claimed six. Its filings say otherwise.
The more they deployed, the less each one was used — Why the scooter startups all went bankrupt0.02.55.07.56.02017Launch2.82018Year 12.52019Year 21.32020Year 31.62021IPO1.32022Year 5Modern MBA
View data
Value
2017 Launch6.0
2018 Year 12.8
2019 Year 22.5
2020 Year 31.3
2021 IPO1.6
2022 Year 51.3
07Total rides is the number that hides itRides taken per year, in millions. It grows with the fleet no matter how badly each vehicle performs.
Total rides is the number that hides it — Why the scooter startups all went bankrupt0M20M40M60M10M2017Launch24M2018Year 140M2019Year 218M2020Year 340M2021IPO47M2022Year 5Modern MBA
View data
Value
2017 Launch10M
2018 Year 124M
2019 Year 240M
2020 Year 318M
2021 IPO40M
2022 Year 547M
08Five years, and the losses grewBird revenue and operating income in millions of dollars.
Five years, and the losses grew — Why the scooter startups all went bankrupt−$600M−$400M−$200M$0M$200M$400M$58M−$369M2018Year 1$141M−$385M2019Year 2$75M−$208M2020Year 3$173M−$235M2021IPO$231M−$471M2022Year 5REVENUEOPERATING INCOMEModern MBA
View data
RevenueOperating income
2018 Year 1$58M−$369M
2019 Year 2$141M−$385M
2020 Year 3$75M−$208M
2021 IPO$173M−$235M
2022 Year 5$231M−$471M
09Even the number they picked themselves looks badAdjusted EBITDA against actual operating income, in millions of dollars. Adjusted EBITDA is the metric a company defines for itself.
Even the number they picked themselves looks bad — Why the scooter startups all went bankrupt−$600M−$400M−$200M$0M−$194M−$369M2018Year 1−$229M−$385M2019Year 2−$182M−$208M2020Year 3−$81M−$235M2021IPO−$62M−$471M2022Year 5ADJUSTED EBITDAOPERATING INCOMEModern MBA
View data
Adjusted EBITDAOperating income
2018 Year 1−$194M−$369M
2019 Year 2−$229M−$385M
2020 Year 3−$182M−$208M
2021 IPO−$81M−$235M
2022 Year 5−$62M−$471M
10The price of a ride doubled and it changed nothingAverage cost of a Bird trip. A 90% increase in five years, into deeper losses.
The price of a ride doubled and it changed nothing — Why the scooter startups all went bankrupt$0.00$2.50$5.00$7.50$2.922017Launch$3.302018Year 1$4.042019Year 2$6.302020Year 3$6.012021IPO$6.312022Year 5Modern MBA
View data
Value
2017 Launch$2.92
2018 Year 1$3.30
2019 Year 2$4.04
2020 Year 3$6.30
2021 IPO$6.01
2022 Year 5$6.31
11Because the customer never paid itAverage cost of a trip split between what the rider paid and what Bird covered. Riders have paid about 80% of the real price since 2018.
Because the customer never paid it — Why the scooter startups all went bankrupt$0.00$2.50$5.00$7.50$2.47$0.832018Year 1$3.49$0.552019Year 2$4.10$2.202020Year 3$4.30$1.702021IPO$4.97$1.342022Year 5PAID BY CUSTOMERCOVERED BY BIRDModern MBA
View data
Paid by customerCovered by Bird
2018 Year 1$2.47$0.83
2019 Year 2$3.49$0.55
2020 Year 3$4.10$2.20
2021 IPO$4.30$1.70
2022 Year 5$4.97$1.34
12Maintenance alone exceeded revenue for three yearsAverage revenue grossed per scooter against maintenance cost per scooter, in dollars.
Maintenance alone exceeded revenue for three years — Why the scooter startups all went bankrupt−$8,000−$6,000−$4,000−$2,000$0$2,000$4,000$2,521−$6,7592018Year 1$3,218−$3,5322019Year 2$1,995−$1,8482020Year 3$2,522−$1,5042021IPO$2,338−$1,0782022Year 5REVENUE PER SCOOTERMAINTENANCE PER SCOOTERModern MBA
View data
Revenue per scooterMaintenance per scooter
2018 Year 1$2,521−$6,759
2019 Year 2$3,218−$3,532
2020 Year 3$1,995−$1,848
2021 IPO$2,522−$1,504
2022 Year 5$2,338−$1,078
13And the vehicle itself cost most of the restAverage revenue per scooter against the cost of the scooter. Generics ran $550 and broke; in-house hardware ran up to $2,479.
And the vehicle itself cost most of the rest — Why the scooter startups all went bankrupt−$4,000−$2,000$0$2,000$4,000$2,521−$5502018Year 1$3,218−$2,3312019Year 2$1,995−$2,4792020Year 3$2,522−$1,3572021IPO$2,338−$1,3582022Year 5REVENUE PER SCOOTERCOST OF SCOOTERModern MBA
View data
Revenue per scooterCost of scooter
2018 Year 1$2,521−$550
2019 Year 2$3,218−$2,331
2020 Year 3$1,995−$2,479
2021 IPO$2,522−$1,357
2022 Year 5$2,338−$1,358
14The hardware never lasted as long as the model saidBest-case projected lifespan against realistic lifespan by model, in months. Bird spent $40 million a year on R&D to get here.
The hardware never lasted as long as the model said — Why the scooter startups all went bankrupt0 mo20 mo40 mo60 mo24 mo12 moBirdZero201828 mo14 moBirdOne201936 mo18 moBirdTwo201948 mo24 moBirdThree202148 mo24 moBird S2021PROJECTEDREALISTICModern MBA
View data
ProjectedRealistic
BirdZero 201824 mo12 mo
BirdOne 201928 mo14 mo
BirdTwo 201936 mo18 mo
BirdThree 202148 mo24 mo
Bird S 202148 mo24 mo
15Franchising moved the cost, it did not remove itGross margin before and after Fleet Managers, who received 100 scooters free, kept 80% of ride revenue and took on all liability for availability.
Franchising moved the cost, it did not remove it — Why the scooter startups all went bankrupt−400%−300%−200%−100%0%100%−365%2018In-house−96%2019Fleet managers8%2020Year 340%2021IPO54%2022Year 5Modern MBA
View data
Value
2018 In-house−365%
2019 Fleet managers−96%
2020 Year 38%
2021 IPO40%
2022 Year 554%
16What a ride would have to costPrice required at current usage to cover unit costs plus a replacement vehicle, against what riders actually pay and an Uber.
What a ride would have to cost — Why the scooter startups all went bankrupt$0.00$5.00$10.00$15.00$20.00$4.972022actual$8.00Break-evenprice$14.59UberModern MBA
View data
Value
2022 actual$4.97
Break-even price$8.00
Uber$14.59
17Even selling the scooters lost moneyGross margin on scooter sales to the public. They could not fold, had no display, were extremely heavy and only worked with the Bird app.
Even selling the scooters lost money — Why the scooter startups all went bankrupt−125%−100%−75%−50%−25%0%25%−101%2019Year 2−55%2020Year 33%2021IPO9%2022Year 5Modern MBA
View data
Value
2019 Year 2−101%
2020 Year 3−55%
2021 IPO3%
2022 Year 59%
18$200 million for $17 million of revenueJUMP revenue and operating income in millions of dollars. Uber bought it in 2018 and sold it to Lime at a loss in 2020.
$200 million for $17 million of revenue — Why the scooter startups all went bankrupt−$300M−$200M−$100M$0M$100M$200M$17M−$50M2018Acquired$119M−$251M2019$35M−$86M2020DivestedREVENUEOPERATING INCOMEModern MBA
View data
RevenueOperating income
2018 Acquired$17M−$50M
2019$119M−$251M
2020 Divested$35M−$86M
19The whole class of 2010s tech, in one yearOperating income in 2022, in millions of dollars. Bird was the smallest loss on this chart and the only one that went bankrupt.
The whole class of 2010s tech, in one year — Why the scooter startups all went bankrupt−$10,000M−$7,500M−$5,000M−$2,500M$0M$2,500M$1,802MAirbnb−$471MBird−$822MShopify−$866MAffirm−$1,027MRobinhood−$1,395MSnap−$8,512MUberModern MBA
View data
Value
Airbnb$1,802M
Bird−$471M
Shopify−$822M
Affirm−$866M
Robinhood−$1,027M
Snap−$1,395M
Uber−$8,512M
01 / 19

Bird revenue, operating income, adjusted EBITDA, fleet size, ride volume, rides per vehicle per day, revenue and maintenance per scooter, vehicle purchase expense, R&D spend, subsidy commitments, scooter sales and gross margins from Bird Global Inc. S-1, 10-K and quarterly filings for the years shown; JUMP revenue and operating income from Uber Technologies filings and acquisition disclosures; funding rounds and private valuations from company announcements; comparative 2022 operating income and revenue multiples for Uber, Airbnb, Shopify, Affirm, Robinhood and Snap from company filings; break-even price is Modern MBA's calculation at reported usage covering unit costs plus a replacement vehicle

Key takeaways

01

The template was already public. Uber lost $3.0B, $4.1B, $3.0B, $8.6B, $4.9B and $3.8B across 2016–2021 to teach people it was cheaper than a taxi, and now an Airbnb reservation averages $164 against $113 in 2017. Scooter startups copied the losing half.

02

Money arrived faster than sense. Lime raised $100M then $405M, Bird $15M then $550M, Spin $8M and Skip $131M — inside two years, for a product with no moat in software or hardware.

03

Bird went from $0 to a $2.0B valuation in one year, $2.5B the next, and $2.3B at IPO — then $0.1B by year five. It was the fastest company in history to $1 billion and it took four years to give all of it back.

04

The fleet scaled like nothing else. 250 scooters in 2017 became 23,000, 44,000, 38,000, 68,600 and 98,800 across 400 cities — an average of roughly 200 to 250 per city, before counting reserves.

05

And the usage collapsed as it grew. Rides per vehicle per day went 6.0, 2.8, 2.5, 1.3, 1.6, 1.3 — falling before COVID and after it. At a 12-minute average trip, the asset is used twelve minutes a day and idle for the rest.

06

That number kills the availability argument. If the average scooter is ridden roughly once a day, people are treating it as a one-way, one-off. More scooters cannot fix that, and neither can looser regulation.

07

Ride volume looked healthy the whole time: 10M, 24M, 40M, 18M, 40M, 47M. Total rides is the metric that hides a broken asset, because it grows with the fleet no matter how badly each unit performs.

08

The revenue never followed. $58M, $141M, $75M, $173M, $231M against operating losses of $369M, $385M, $208M, $235M, $471M. Even Bird's own cherry-picked adjusted EBITDA ran $194M to $62M.

09

Raising the price did not help. The average ride went from $2.92 to $6.31 — a 90% increase in five years — and the losses got bigger, not smaller.

10

Because customers never paid the sticker. Bird subsidized $20M, $22M, $40M, $69M and $62M of trip costs, so the rider paid $2.47 of a $3.30 ride and $4.97 of a $6.31 one. Customers have covered about 80% of the true price since 2018 and Bird ate the rest.

11

Per scooter, the business got worse as it matured. Revenue per vehicle went $2,521, $3,218, $1,995, $2,522, $2,338 against maintenance of $6,759, $3,532, $1,848, $1,504 and $1,078. Maintenance alone exceeded revenue for the first three years.

12

Hardware was a trap in both directions. Generic Chinese scooters cost $550 and broke constantly; in-house models cost $2,331, then $2,479, then settled at $1,357 — and still only lasted 12 to 24 months against the 24 to 48 Bird projected.

13

So Bird invented a franchise. Fleet Managers got 100 scooters at no cost and kept 80% of ride revenue while becoming liable for every part of availability. Gross margin went from 365% to 96%, then 8%, 40% and 54% — the cost moved, it did not disappear.

14

The break-even is the whole case. At current usage, covering unit costs plus one replacement vehicle requires a rider to pay $8.00 a trip against the $4.97 they pay. That is Uber pricing at $14.59, for a scooter.

15

Even the salvage value was fake. Bird spent $102M, $12M, $215M and $89M a year buying vehicles and $40M a year on R&D, then sold scooters to the public at gross margins of 101%, 55%, 3% and 9% — they could not be folded, had no display, were extremely heavy and only worked with the Bird app.

16

The buyers learned it too. Uber paid $200M for JUMP on $17M of revenue, watched it lose $251M on $119M, sold it to Lime at a loss and landfilled the bikes. Ford paid $100M for Spin and shut it down in 2022. The founders and VCs cashed out at IPO and now work at AI startups.

Common questions

Why did Bird go bankrupt?

Because the asset never paid for itself. The average Bird scooter grossed around $2,000 to $2,500 a year, cost between $550 and $2,479 to buy depending on the model, and cost roughly $1,000 to $1,500 a year to maintain — and lasted 12 to 24 months in practice against the 24 to 48 the company projected. Usage fell from 2.8 rides per vehicle per day to 1.3, which at a 12-minute average trip means the scooter is earning for twelve minutes and idle for the rest of the day while its battery drains and someone has to collect and recharge it. Bird raised prices 90% over five years and lost more money, not less. It was delisted from Nasdaq and filed for bankruptcy in late 2023.

Why did scooter companies lose money on every ride?

Because the price was set below cost on purpose, and the cost base never came down. Bird committed $20 million to $69 million a year in discounts and promotions, so the average customer paid about 80% of what a trip actually cost — $2.47 of a $3.30 ride in 2018, $4.97 of a $6.31 ride in 2022. The strategy was Uber's: teach consumers the product is cheaper than the bus, then raise prices later. But Bird was never able to commit to either lever. It raised prices to reduce losses while staying too afraid to cut the subsidies, so it got neither the volume nor the margin. To actually break even at reported usage, a rider would have to pay $8.00 a trip.

How many times a day is a shared scooter actually used?

About once. Bird publicly claimed six rides per scooter per day; its audited filings show 6.0 in the launch year, then 2.8, 2.5, 1.3, 1.6 and 1.3. Anything above two is an outlier. With an average trip of 12 minutes, that means a scooter generates revenue for roughly twelve minutes out of every twenty-four hours. It also tells you the behavior: people use these as opportunistic one-ways rather than routine round trips, which is why deploying more scooters never increased usage per scooter — it just spread the same demand across more assets.

What was the Fleet Manager model?

Franchising, with the depreciation pushed onto the franchisee. Bird started with in-house operations, moved to gig workers who collected and charged scooters overnight for bonuses, then in 2019 appointed Fleet Managers in each city — handed them 100 scooters at no cost and kept 20% of every ride while the manager kept 80%. In exchange, the Fleet Manager became financially responsible for availability: if the scooters were not rented, nobody earned, and the vehicles were already sunk cost for Bird. It worked as intended — gross margin went from −365% to 54% — but the underlying cost of picking up, charging and repairing scooters did not go away. It moved to someone with less capital to absorb it.

Why didn't better scooters fix the economics?

Because durability was never the binding constraint. Bird, Lime and Spin all launched on the same generic Chinese scooters at around $550 a unit, which broke constantly under public use, weather and manhandling. Building in-house pushed the cost to $2,331 and then $2,479 before settling around $1,357, with bigger batteries, heavier shock absorbers and the sidewalk-detection sensors cities began demanding for permits. Bird spent $40 million a year on R&D and $100 million a year on replacement vehicles, and its best model still managed 12 to 24 real months. Riders could not tell the models apart. Even at a perfect lifespan, a vehicle earning $2,000 a year against $1,300 of hardware and $1,000 of maintenance does not clear.

Did anyone make money in micromobility?

The people who sold. Uber paid $200 million for JUMP when it had $17 million of revenue, held it three years while it lost $251 million on $119 million of revenue, then sold it to Lime at a loss and put the bikes in landfill. Ford paid $100 million for Spin and shut it down in 2022. The venture capitalists who created the category cashed out at IPO, before Bird fell to a penny stock. The founders and executives of these companies now work at AI startups. Bird itself was taken over by Bird Canada, which has exited cities, moved to bikes in Europe, stopped selling scooters, and is trying to make American metros profitable.

Discussion

  1. The venture playbook was price below cost, deploy without asking, settle with the city afterwards. It built Uber and Airbnb. Why did the same sequence fail here?

  2. Bird claimed six rides per scooter per day; the filings show 2.8, then 2.5, 1.3, 1.6. Who was in a position to check the claim before the money went in, and why did nobody?

  3. Uber paid $200 million for JUMP on $17 million of revenue and Ford paid $100 million for Spin. What were the buyers actually acquiring?

  4. The pitch was disrupting walking — and that is written down, not hindsight. What should a founder or investor conclude when the addressable market is a thing people already do for free?

  5. You are handed a scooter company with two years of runway. Is there a version of this that works, and what does it have to stop doing?

Related case studies