Rideshare Economics 102

[hwhelchel]

What do you want to see covered in the TRIP Report newsletter?

I'd like to see an issue of the TRIP Report that explores the macro view of the rideshare industry.

Areas of coverage could include:

• History of rideshare in 2010s.
• Regional differences in the evolution of rideshare (North America vs LATAM vs Asia vs Europe)
• Expansion of rideshare derivatives and the definition of the broader mobility category (food delivery, scooters, charter, etc)
• History of Uber, their fundraising, and what those funds were and are currently used for.
• Prediction by McKinsey and other leading industry analysts on the growth of rideshare and how mobility will involve through 2030.
• How those predictions can and should be modified in light of addition of TRIP to the industry.

Various stakeholders including the leadership team at DEC, current and prospective investors, as well as community members are interested have a document that articulates our community's vision of the broader landscape of rideshare, how TRIP fits within it, and where we see rideshare going.

Are you interested in writing this newsletter yourself?

I would gladly support and advise anyone interested in contributing this issue.

[hwhelchel]

Here is some content that needs polishing but can be used in whole or in part to create this newsletter issue.

Note that some of the Uber vs TRIP Rideshare Economics breakdown charts may be from an older version of that chart. Make sure to reference the published version of that chart here and use those numbers instead.

---

Introduction

The current rideshare market is dominated by centralized platforms that control pricing, terms of service, and revenue distribution. As a result drivers earn less and riders pay more than otherwise if there weren’t centralized players exercising their monopoly power over the network.

In a decentralized rideshare market, there are no centralized platforms that control pricing, terms of service, or revenue distribution. Instead an open protocol enables local rideshare companies to emerge, compete, and form a global network.

The open nature of the protocol leads to greater competition, driving down the take rate of the rideshare operators, leading to drivers earning more and riders paying less.

"TRIP - The Rideshare Protocol" presents a decentralized alternative, proposing a blockchain-based, open protocol that enables local rideshare companies to form a global network. This model aims to reduce costs for riders, offer fairer compensation for drivers, and keep more revenue within local economies, while promoting self-governance through distributed consensus technology.

The Rideshare Landscape: An Overview

Over the decade spanning 2010 to 2020, ridesharing has evolved from a nascent idea to a global industry with estimated revenues in excess of $100B.

In North America, Uber and Lyft are the dominant two ridesharing companies, with Uber taking up 74% of U.S sales. Uber and Lyft (originated from Zimride) the earliest ridesharing companies as we know today, were founded around the 2007-2009 period. However, similar companies began to be founded outside North America a few years later in 2011-2013.

Company	Region	Year Founded	Other Comments
Didi	China	2012	Uber has ~12% ownership [20] after Uber China was acquired by Didi in 2016
Bolt (Taxify)	Europe	2013	N/A
Careem	Middle East	2012	Uber acquired Careem in 2019
Grab	South East Asia	2012	Uber has ~14% ownership [19]
Yandex	Russia	2011	N/A

2017: https://qz.com/1320449/this-simple-formula-could-increase-pay-for-uber-drivers-in-new-york | | Lyft | 2017: https://qz.com/1320449/this-simple-formula-could-increase-pay-for-uber-drivers-in-new-york | | Juno | 2017: https://qz.com/1320449/this-simple-formula-could-increase-pay-for-uber-drivers-in-new-york | | Via (shutdown 2021) | 2017: https://qz.com/1320449/this-simple-formula-could-increase-pay-for-uber-drivers-in-new-york | | Misc. | 2017: https://www.sfcta.org/sites/default/files/2019-02/TNCs_Today_112917_0.pdf | | Taxi | 2015: https://www.autoinsurance.com/research/rideshare-statistics/ | | Anecdotal (individual) | https://youtu.be/6y8Op9_FL3I?si=VtTv4Hd-6XjBZdbh&t=1063 |

City Launch Criteria

Agent-Based-Modeling (ABM) of Hailing Mobility

Agent-Based-Modeling computational approaches are used to simulate the actions and interactions of agents (in the current context, riders and drivers) in order to assess:

1. City-to-city variations in ride hailing metrics
2. Future: Response to incentive structures

TODO:

• Provide visualizations that tell a good story on the work. Good story means:
• Show that we can generalize to different maps using the OSM framework
• (FURTHER FUTURE:) Explanation videos detailing how we leverage simulations to strategize city launches.

WIP: Looking at Uber and Lyft Annual Financials (K-10)

1.1.1 US

According to Bloomberg Second Measure, in October 2023, Uber accounted for 74% of observed U.S. rideshare spending, with the other primary competitor Lyft taking a large portion of the remaining.

Problem is if you only have effectively two competitors it's possible to still have implicit price coordination.

The U.S. Rideshare Industry: Uber vs. Lyft - Bloomberg Second Measure

1.2 Mobility Unit Economics

We estimate that Uber’s existing take rate as of 2023 to be as high as 44%, assuming a cost breakeven scenario.

An estimate of Uber’s unit economics at a ride level. Estimates were extracted from Uber’s Annual Financials in addition external analysts’ work. For a detailed description of these categories and the approach through which their values were obtained, refer to Appendix A1:

Uber’s take rate at 44% can be decomposed into 29.5% variable costs and 14.9% fixed costs. In relative terms, Uber is spending 66% of its revenue on variable costs and 34% on fixed costs. Based on our estimates, Uber’s take rate has been increasing from 34% to 44% from 2019 to 2023, which corroborates external findings.

Variable Costs

A large portion of Uber’s spend consists of variable costs for rideshare operations, more notedly, user acquisition (driver and rider). This spending is combined in the chart above under the “Incentives & Advertising” category which constitutes a significant 20.2% of a ride fare. Reducing or eliminating this category provides margin for drivers to be paid more and riders to pay less.

The other variable costs (e.g., Support, Insurance, Credit Card (CC) Processing Fee) are relatively constant in their relative percentages to the ride fare.

Fixed Costs

At Uber’s scale, fixed costs are the cascading costs of a centralized organization. The following section describes these components:

A substantial 5% is categorized as "Cost of Revenue (Overhead)". These are estimated from balances unaccounted for

Remaining competitive in a centralized manner requires sustained growth of R&D (hiring and resources), which also increases General Administrative costs. Annual restructuring contributes to small costs. Together, these organizational costs take away 4.3% from each ride.

Housing employees also incurs overhead costs like corporate office leases, real estate, utilities, and maintenance, categorized as Variable Lease. Additionally, there are asset-related costs:

• Operating lease for Uber's vehicles (including light vehicles)
• Amortization and depreciation of physical and intangible assets (e.g., computer equipment, furniture, land, software, etc.)
• Asset impairments of mentioned assets

This means that Uber spends 3.8% of ride fare on maintaining its assets.

Uber's centralized approach to ridesharing across different regulations makes it vulnerable to legal liabilities. Therefore, part of their expenses goes to a "legal, tax, and regulatory reserve", which constituted 0.2% of gross bookings in 2023. This value fluctuates yearly, as reached as high as 1.4% in 2022. Since Uber controls the take rate algorithmically, any regulatory settlement costs can potentially be passed on to drivers.

## Introduction The current rideshare market is dominated by centralized platforms that control pricing, terms of service, and revenue distribution. As a result drivers earn less and riders pay more than otherwise if there weren’t centralized players exercising their monopoly power over the network. In a decentralized rideshare market, there are no centralized platforms that control pricing, terms of service, or revenue distribution. Instead an open protocol enables local rideshare companies to emerge, compete, and form a global network. The open nature of the protocol leads to greater competition, driving down the take rate of the rideshare operators, leading to drivers earning more and riders paying less. "TRIP - The Rideshare Protocol" presents a decentralized alternative, proposing a blockchain-based, open protocol that enables local rideshare companies to form a global network. This model aims to reduce costs for riders, offer fairer compensation for drivers, and keep more revenue within local economies, while promoting self-governance through distributed consensus technology. ## **The Rideshare Landscape: An Overview** Over the decade spanning 2010 to 2020, ridesharing has evolved from a nascent idea to a global industry with estimated revenues in excess of $100B. In North America, Uber and Lyft are the dominant two ridesharing companies, with Uber taking up 74% of U.S sales. Uber and Lyft (originated from Zimride) the earliest ridesharing companies as we know today, were founded around the 2007-2009 period. However, similar companies began to be founded outside North America a few years later in 2011-2013. | Company | Region | Year Founded | Other Comments | | --- | --- | --- | --- | | Didi | China | 2012 | Uber has ~12% ownership [20] after Uber China was acquired by Didi in 2016 | | Bolt (Taxify) | Europe | 2013 | N/A | | Careem | Middle East | 2012 | Uber acquired Careem in 2019 | | Grab | South East Asia | 2012 | Uber has ~14% ownership [19] | | Yandex | Russia | 2011 | N/A | Despite Uber’s original goal of aggressive growth to monopolize global market share, one observes at the global scale that company dominance varies by region, hinting at the diverse regulatory landscape that a ridesharing company must navigate in order to penetrate new markets at the international level.  [Team of rivals: The startups Uber competes against around the world | Mashable](https://mashable.com/article/uber-global-rivals-didi) Over the next decade, as the definitions of ridesharing continue to expand to potentially include carpooling, shuttle services, shared autonomous vehicles, the global ridesharing market size has been projected sustain growth to reach $300-$860 billion by 2030 [1-3]. The figures below are obtained from McKinsey and Oliver Wyman Forum independently. Observe the numbers specific to “Hailed Mobility” or “Ride-hailing” in both diagrams.   The other value pools seen here (e.g., micromobility, aerial mobility, car rentals) are starting to be grouped alongside hailed mobility under the terms “Mobility-as-a-Service” or “Shared Mobility”. These value pools present themselves as growth alternatives in addition to the conventional ridesharing value pool, though not focal in the current report. ## **Economics of Classical Rideshare: A Deep Dive** Intended to supplant the taxi industry, Uber’s strategy was to sacrifice short-term profitability in exchange for maximizing long-term market share growth. This meant aggressive usage of capital to: - attract riders with subsidized ride fares lower than local competitors - attract drivers with additional driver incentives Uber’s monopolistic growth strategy was executed in the hopes of being able to make up its losses by adjusting its pricing schemes when substantial market share was eventually achieved. One major problem is that users of ridesharing apps are usually brand agnostic, and one may argue that Uber’s recent offerings in Uber One and Uber Ads is an attempt to build consumer brand loyalty. As highlighted from this [[article](https://innovestor.substack.com/p/uber-deep-dive?nthPub=111)](https://innovestor.substack.com/p/uber-deep-dive?nthPub=111): > *Part of Uber’s problem here is the fact that users of ride-sharing apps don’t really tend to have much loyalty or care about which service they are using**.*** > > - *Only 34% of **customers** use both Uber and Lyft* > - *2/3 of **drivers** drive for both apps* > - *68% of drivers stop driving within 6 months* > - *Churn levels are reported to be as high as 96% after 1 year* In other words, Uber’s strategy towards maximizing user market share is hindered cyclically by: 1. The inability to profitably compensate drivers in the short term 2. Lack of consistent profitability on the drivers’ end leads to drivers churning 3. Needing to compensate resulting driver churn via cash incentives in the medium term This perpetual cycle of having to consistently incentivize poorly compensated drivers while growing is a major contributor towards sustained losses over the past decade, in addition to overly aggressive expansion attempts at the global level and several other speculative investments.  Uber’s cumulative losses visualized on a quarterly basis [18] Uber reported its first profitable quarter in 2023 Q2 with an excess of $326M. However, the reported profitable quarter contains “contributions from an alleged $386 million second quarter gain in the value of untradable securities they hold in companies like Didi, Grab, and Aurora that have nothing to do with their ongoing operations.” [17]. Note the relative size of $386 million against the reported operating profit of $326 million. Excluding this contribution of untradeable stocks, it is likely that Uber is breaking at best. Assuming a break-even scenario, it is estimated that Uber’s take-rate in 2023 is at 44.4% which corroborates findings from Len Sherman’s analysis based on data from YipitData and Gridwise that places Uber’s average take-rate at 40% [11]. In contrast, Uber’s 2023 quarterly financials currently reports a take-rate or “revenue margin” of 28% for 2023 Q3.

💡 The take-rate in this report is defined as being that perceived by the driver and is consistent with how Len Sherman has defined the take-rate. $$ \text{Uber Take Rate} = \frac{\text{Price - Pay}}{\text{Price}} $$ Contrarily, Uber currently reports a take-rate or “revenue margin” of 28% for 2023 Q3. Indeed, this discrepancy between driver perceived take-rate vs. Uber’s take-rate has existed since 2019, where Uber reported their take-rate at 21% [TODO] but an independent survey saw that take-rates for Uber averaged at 35% while Lyft averaged 38% [9]

To compare the centralized and decentralized business models, the unit economics of ridesharing is considered. More specifically, assuming breakeven conditions, one can investigate the cost breakdown of revenue from a single ride. An advantageous business model would highlight a higher take-rate for the driver. A direct side-by-side comparison of the models is provided below, illustrating driver take-rates in Uber (56%) vs. TRIP (85%).  Internally sourced: A comparison of a ride fare breakdown in Uber against a ride in TRIP Protocol. Refer to [Appendix A1](https://www.notion.so/Rideshare-Economics-Primer-Paul-s-Edit-Thad-Eugene-s-Work-Notes-e67b90ae714c47c69e216ed78887c00d?pvs=21) for more details on the breakdown. A decentralized model under the TRIP Protocol offers two key advantages over that of a centralized model (Uber as an example):

💡 1. Reduces ***Incentives & Advertising*** costs 2. Eliminates certain ***fixed costs*** that come with centralization

which is demonstrated in the TRIP portion of the chart where non-applicable costs are greyed out. To fully appreciate this comparison, lets make sense of the cost categories listed on Uber’s side. An estimate of Uber’s unit economics at a ride level. Estimates were extracted from Uber’s Annual Financials in addition external analysts’ work. For a detailed description of these categories and the approach through which their values were obtained, refer to [Appendix A1](https://www.notion.so/Rideshare-Economics-Primer-Paul-s-Edit-Thad-Eugene-s-Work-Notes-e67b90ae714c47c69e216ed78887c00d?pvs=21). Of Uber’s 44% take-rate, a substantial 66% goes towards variable costs for rideshare operations. Variable costs are costs that scale directly with the amount of rides, like the credit card processing fee per ride, customer support required to address inquiries, and driver insurance. **20.2% of each ride is spent on customer acquisition** (driver and rider) under the “***Incentives & Advertising***” category. Note that there’s a circular effect not highlighted in the graph, in which the absolute rider fare is currently subsidized by said incentives and drivers are also incentivized separately to provide rides. In the absence of these subsidies, the riders would pay more and drivers would pay less.

💡 At Uber’s scale, fixed costs can be interpreted as cascading costs of a centralized organization, which may not immediately apply to a local decentralized operator.

**A substantial 5% is spent on "Cost of Revenue (Overhead)"**. These are estimated from excess balance from the Cost of Revenue segment in the balance sheet after subtracting known contributing costs for data servers hosting, insurance, credit card processing fees, legal settlements. **Organizational costs take away 4.3% from each ride**. Remaining competitive in a centralized manner requires sustained growth in R&D (hiring and resources), which consequently increases General Administrative costs. Annual restructuring of an increased organization also contributes to incrementally. **Asset (physical and non-tangible) maintenance or overhead takes away 3.8% of ride fare**. Housing employees also incurs overhead costs like corporate office leases, real estate, utilities, and maintenance, categorized as Variable Lease. Other asset-related costs are also included: - Operating lease for Uber's vehicles (including light vehicles) - Amortization and depreciation of physical and intangible assets (e.g., computer equipment, furniture, land, software, etc.) - Asset impairments of mentioned assets **Legal settlements took away 0.2% from gross bookings in 2023**. Uber's centralized approach to ridesharing across different regulations makes it vulnerable to legal liabilities. This value fluctuates yearly and reached as high as 1.4% in 2022. Since Uber controls the take rate algorithmically, any regulatory settlement costs can potentially be passed on to drivers. A lot of these fixed costs can be substantially reduced or eliminated in a decentralized ridesharing system. Under the TRIP Protocol, Uber’s role is equivalent to an Operator who is operating at a very large, centralized scale. Under the TRIP Protocol, it is expected that many local, decentralized Operators would participate in the network. A decentralized Operator operating at a smaller scale does not require a substantial workforce to keep its operations running. The overhead costs associated with asset management is also greatly reduced. More importantly, operating under the TRIP Protocol allows Operators to leverage permissionless marketing conducted by the network participants, and eliminate the hard requirement of having to market their businesses. Although Operators can market themselves to gain a competitive edge of drivers connecting to their service, the act of securing the riders and drivers to the network as a whole propagates through permissionless marketing. Overall, the reduction of these costs under the TRIP Protocol are reflected in the greyed out categories chart below:  In the chart above, Operators have a 9.5% take-rate based on a break-even assumption, in order to have an apples-to-apples comparison with Uber. Here, an assumed 3.1% is spent on customer support, 3.6% on insurance, 2% on credit card processing fee and 0.8% on hosting infrastructure, where values are assuming a certain level of scaling. Additionally, an assumed 0.5% of the take-rate goes to Client App providers. In practice, the Operator and Client App take-rates will likely vary from provider-to-provider, as they compete for drivers and a corresponding trip completion rate from said drivers. The only encoded take-rate is 5% for the TRIP Protocol, which will be channeled to a community treasury that will be decentralized incrementally. # References: 1. [[Heineke, K., Kloss, B., Mertens-von Rüden, A., Möller, T., & Wiemuth, C. (2023). Shared mobility: Sustainable cities, shared destinies. McKinsey Center for Future Mobility.](https://www.mckinsey.com/industries/automotive-and-assembly/our-insights/shared-mobility-sustainable-cities-shared-destinies#/download/%2F~%2Fmedia%2Fmckinsey%2Findustries%2Fautomotive%20and%20assembly%2Four%20insights%2Fshared%20mobility%20sustainable%20cities%20shared%20destinies%2Fshared-mobility-sustainable-cities-shared-destinies-final.pdf%3FshouldIndex%3Dfalse)](https://www.mckinsey.com/industries/automotive-and-assembly/our-insights/shared-mobility-sustainable-cities-shared-destinies#/download/%2F~%2Fmedia%2Fmckinsey%2Findustries%2Fautomotive%20and%20assembly%2Four%20insights%2Fshared%20mobility%20sustainable%20cities%20shared%20destinies%2Fshared-mobility-sustainable-cities-shared-destinies-final.pdf%3FshouldIndex%3Dfalse) 2. [[Goldman Sachs Global Investment Research. (2019, June 4). The Future of Mobility.](https://www.goldmansachs.com/intelligence/pages/gs-research/future-of-mobility/report.pdf)](https://www.goldmansachs.com/intelligence/pages/gs-research/future-of-mobility/report.pdf) 3. [[Value Pool Report. Oliver Wyman Forum. (November 2022). The Future of Mobility 2030.](https://www.oliverwymanforum.com/content/dam/oliver-wyman/ow-forum/mobility/2022/Value%20Pool%20Report.pdf)](https://www.oliverwymanforum.com/content/dam/oliver-wyman/ow-forum/mobility/2022/Value%20Pool%20Report.pdf) 4. [[Kaczmarski, M. (2023, November 13). Uber vs. Lyft: Who’s tops in the battle of U.S. rideshare companies. Bloomberg Second Measure.](https://secondmeasure.com/datapoints/rideshare-industry-overview/#:~:text=Looking%20at%20the%20rideshare%20competitors,July%20and%20August%20of%202023)](https://secondmeasure.com/datapoints/rideshare-industry-overview/#:~:text=Looking%20at%20the%20rideshare%20competitors,July%20and%20August%20of%202023) 5. [[Rideshare Statistics for 2023 | AutoInsurance.com](https://www.autoinsurance.com/research/rideshare-statistics/)](https://www.autoinsurance.com/research/rideshare-statistics/) 6. [[Uber drivers in New York could earn more with this simple formula (qz.com)](https://qz.com/1320449/this-simple-formula-could-increase-pay-for-uber-drivers-in-new-york)](https://qz.com/1320449/this-simple-formula-could-increase-pay-for-uber-drivers-in-new-york) 7. [Boeing, G. Urban spatial order: street network orientation, configuration, and entropy. *Appl Netw Sci* **4**, 67 (2019). https://doi.org/10.1007/s41109-019-0189-1](https://appliednetsci.springeropen.com/articles/10.1007/s41109-019-0189-1) 8. [[Roundtable Discussion:: Rideshare & Gig Work Utilization Rate with @Therideshareguys - YouTube](https://www.youtube.com/watch?v=6y8Op9_FL3I&t=1063s)](https://www.youtube.com/watch?v=6y8Op9_FL3I&t=1063s) 9. [[Uber And Lyft Take A Lot More From Drivers Than They Say (jalopnik.com)](https://jalopnik.com/uber-and-lyft-take-a-lot-more-from-drivers-than-they-sa-1837450373)](https://jalopnik.com/uber-and-lyft-take-a-lot-more-from-drivers-than-they-sa-1837450373) 10. [[Uber Deep-Dive - Innovestor Deep Dives (substack.com)](https://innovestor.substack.com/p/uber-deep-dive)](https://innovestor.substack.com/p/uber-deep-dive) 11. [[Uber’s CEO Hides Driver Pay Cuts To Boost Profits (forbes.com)](https://www.forbes.com/sites/lensherman/2023/12/11/ubers-ceo-hides-driver-pay-cuts-to-boost-profits/?sh=4b79c39f74fd)](https://www.forbes.com/sites/lensherman/2023/12/11/ubers-ceo-hides-driver-pay-cuts-to-boost-profits/?sh=4b79c39f74fd) 12. [[Insurance: The Primary Barrier To Uber's Profitability (NYSE:UBER) | Seeking Alpha](https://seekingalpha.com/article/4293755-insurance-primary-barrier-to-ubers-profitability)](https://seekingalpha.com/article/4293755-insurance-primary-barrier-to-ubers-profitability) 13. [[Lyft - Insurance And The Path To Profitability (NASDAQ:LYFT) | Seeking Alpha](https://seekingalpha.com/article/4249415-lyft-insurance-and-path-to-profitability)](https://seekingalpha.com/article/4249415-lyft-insurance-and-path-to-profitability) 14. [[Uber’s Balance Sheet Says… We’re Not A Tech Company | by Asher M | Medium](https://medium.com/@AsherOfLA/ubers-balance-sheet-says-we-re-not-a-tech-company-e3b6a9561d14)](https://medium.com/@AsherOfLA/ubers-balance-sheet-says-we-re-not-a-tech-company-e3b6a9561d14) 15. [[Uber Used 50 Dutch Shell Companies to Avoid Taxes on $5.8 Billion Revenue: Report (businessinsider.com)](https://www.businessinsider.com/uber-tax-avoidance-50-dutch-shell-companies-5-billion-revenue-2021-5#:~:text=Uber%20used%20around%2050%20Dutch,Champions%20League%20of%20tax%20avoidance.%22)](https://www.businessinsider.com/uber-tax-avoidance-50-dutch-shell-companies-5-billion-revenue-2021-5#:~:text=Uber%20used%20around%2050%20Dutch,Champions%20League%20of%20tax%20avoidance.%22) 16. [[Uber's Elaborate Tax Scheme Explained | Fortune](https://fortune.com/2015/10/22/uber-tax-shell/)](https://fortune.com/2015/10/22/uber-tax-shell/) 17. [[Hubert Horan: Can Uber Ever Deliver? Part Thirty-Three: Uber Isn’t Really Profitable Yet But is Getting Closer; The Antitrust Case Against Uber | naked capitalism](https://www.nakedcapitalism.com/2023/08/hubert-horan-can-uber-ever-deliver-part-thirty-three-uber-isnt-really-profitable-yet-but-is-getting-closer-the-antitrust-case-against-uber.html)](https://www.nakedcapitalism.com/2023/08/hubert-horan-can-uber-ever-deliver-part-thirty-three-uber-isnt-really-profitable-yet-but-is-getting-closer-the-antitrust-case-against-uber.html) 18. [[Money-driven: Uber's operations are finally profitable (chartr.co)](https://www.chartr.co/stories/2023-08-02-1-ubers-operations-finally-profitable)](https://www.chartr.co/stories/2023-08-02-1-ubers-operations-finally-profitable) 19. [[With 31% ownership of the shares, Grab Holdings Limited (NASDAQ:GRAB) is heavily dominated by institutional owners (yahoo.com)](https://finance.yahoo.com/news/31-ownership-shares-grab-holdings-101914053.html)](https://finance.yahoo.com/news/31-ownership-shares-grab-holdings-101914053.html) 20. [Uber lost $2 billion in Didi stake this week on China crackdown threat ([cnbc.com](http://cnbc.com/))](https://www.cnbc.com/2021/07/23/uber-lost-2-billion-in-didi-stake-this-week-on-china-crackdown-threat.html#:~:text=Uber%20owns%20about%2012%25%20of%20Didi%2C%20making%20it,Didi%20in%20exchange%20for%20equity%20in%20its%20rival.) # Appendix ## A1: Description of Uber’s Expense Breakdown Categories ### Variable Costs | | Description | Comments | | --- | --- | --- | | Incentives & Advertising | Lumped category containing Driver Incentives, Rider Incentives, Advertising expenses and Sales & Marketing wages | Historically, Uber has grouped incentives under the Cost of Revenue segment in their financials. This category includes an estimate from that segment. | | Support | Compensation costs (excluding stock-based) for operations and customer support | | | Insurance | Insurance coverage for drivers (Mobility and Delivery segments) | Assumed 3.6% based on external financial analysis modeling from JP Morgan. | | CC Processing Fee | Credit Card Processing Fees | Assumed 2% based of internal negotiations with vendors | ### Fixed Costs | | Description | Comments | | --- | --- | --- | | General Administrative Wages | Wages for “executive management and administrative employees, including finance and accounting, human resources, policy and communications, legal” | | | R&D Wages | Includes compensation for research, engineering | | | Hosting Infrastructure | Includes costs for hosting the core platform | Assumed hosting costs of 1.6%. Value obtained from financial analyst. | | Depreciation & Amortization | “primarily consist of depreciation on buildings, site improvements, computer and network equipment, software, leasehold improvements, furniture and fixtures, and amortization of intangible assets” | | | Legal, Tax, & Regulatory | “We are involved in legal proceedings, claims, and regulatory, indirect tax examinations, or government inquiries and investigations that may arise in the ordinary course of business” - from Loss Contingencies of Uber K-10 Financial Statements | | | Operating Lease | Vehicles (including light vehicles) leasing | Gross Booking Percentage extrapolation from 2022. | | Variable Lease | “Variable lease payments may include costs such as common area maintenance, utilities, real estate taxes or other costs” ”Our leases primarily include corporate offices, data centers, and servers.” | Gross Booking Percentage extrapolation from 2022. | | Asset Impairment | Asset impairment of “Intangible Assets, Property and equipxment, Operating lease right-of-use-assets” as observed under “Impairment of Definite-Lived Intangible and Long-Lived Assets” | | | Restructuring Changes | “Costs associated with management-approved restructuring activities, including reductions in headcount, exiting a market or consolidation of facilities are recognized when they are incurred and may include employee termination benefits, impairment of long-lived assets (including impairment of operating lease right-of-use assets), contract termination costs and accelerated lease cost for right-of-use assets that ceased to be used” | | | Cost of Revenue (Overhead) | Uber defines the Cost of Revenue as: “primarily consists of certain insurance costs related to our Mobility and Delivery offerings, credit card processing fees, bank fees, data center and networking expenses, mobile device and service costs, costs incurred with Carriers for Uber Freight transportation services, amounts related to fare chargebacks and other credit card losses as well as costs incurred for certain Mobility and Delivery transactions where we are primarily responsible for Mobility or Delivery services and pay Drivers and Couriers for services” Due to “business model impacts” leading to new accounting methods, certain geo-specific driver payouts started to be lumped under Cost of Revenue, substantially inflating the Cost of Revenue compared to previous years. It is likely that this category is a mixture of real expense and the consideration of driver payouts as well. | After removing costs associated with: - Carriers for Uber Freight - Operating Lease - Hosting - Insurance - Credit Card Processing Fees - Legal, Tax & Regulatory This category is estimated by subtracting a projection of incentives spend away from the unaccounted balance. | # Support Content (extra) # Section 2: Launching in a City In launching a two-sided marketplace in a particular city, core metrics towards customer satisfaction levels have to be evaluated on both the supply and demand side. A summary of target core metrics is provided here: The above target metrics are subsequently used to calculate the number of required referred riders and drivers for a particular city prior to launch. These threshold numbers are derived with the goal of forming an atomic network with critical mass for continued growth via permissionless marketing. ## Demand Side Metrics (Riders) On the demand side, **rider wait times** forms the core metric. Wait times can be further broken down into two categories: 1. Hailing wait time - The amount of time taken for a local operator to match a rider with a driver, measured from the start of the hailing process. 2. Pickup wait time - The amount of time taken for a driver to pickup a rider, measured from the start of a mutually accepted driver-rider match. ## Supply Side Metrics (Drivers) On the supply side, **driver utilization** is the core metric, defined as the amount of time a driver is busy in a single hour. This metric also highly correlates to driver earnings on the hour, which may vary geographically. Using utilization time Current utilization rates at Uber and Lyft usually converge to the range 40-60% over time `TODO: Need Citation`. Higher utilization rates are good for drivers, but occurs when demand exceeds that of the supply, and could be transient (vs. long-term) behavior. This conversely results in lower rider hail reliability (e.g. wait times, hail success). A collection of driver utilization rates from various studies is listed below: | Company | Utilization Rate | | --- | --- | | Uber | 2015: https://www.autoinsurance.com/research/rideshare-statistics/ 2017: https://qz.com/1320449/this-simple-formula-could-increase-pay-for-uber-drivers-in-new-york | | Lyft | 2017: https://qz.com/1320449/this-simple-formula-could-increase-pay-for-uber-drivers-in-new-york | | Juno | 2017: https://qz.com/1320449/this-simple-formula-could-increase-pay-for-uber-drivers-in-new-york | | Via (shutdown 2021) | 2017: https://qz.com/1320449/this-simple-formula-could-increase-pay-for-uber-drivers-in-new-york | | Misc. | 2017: https://www.sfcta.org/sites/default/files/2019-02/TNCs_Today_112917_0.pdf | | Taxi | 2015: https://www.autoinsurance.com/research/rideshare-statistics/ | | Anecdotal (individual) | https://youtu.be/6y8Op9_FL3I?si=VtTv4Hd-6XjBZdbh&t=1063 | ## City Launch Criteria ## Agent-Based-Modeling (ABM) of Hailing Mobility Agent-Based-Modeling computational approaches are used to simulate the actions and interactions of agents (in the current context, riders and drivers) in order to assess: 1. City-to-city variations in ride hailing metrics 2. Future: Response to incentive structures `TODO:` - Provide visualizations that tell a good story on the work. Good story means: - Show that we can generalize to different maps using the OSM framework - (FURTHER FUTURE:) Explanation videos detailing how we leverage simulations to strategize city launches. [[WIP: Looking at Uber and Lyft Annual Financials (K-10)](https://www.notion.so/WIP-Looking-at-Uber-and-Lyft-Annual-Financials-K-10-ea2af9a03fda43c8a858922147b8fefd?pvs=21)](https://www.notion.so/WIP-Looking-at-Uber-and-Lyft-Annual-Financials-K-10-ea2af9a03fda43c8a858922147b8fefd?pvs=21) ### 1.1.1 US According to Bloomberg Second Measure, in October 2023, Uber accounted for 74% of observed U.S. rideshare spending, with the other primary competitor Lyft taking a large portion of the remaining. Problem is if you only have effectively two competitors it's possible to still have implicit price coordination.  [The U.S. Rideshare Industry: Uber vs. Lyft - Bloomberg Second Measure](https://secondmeasure.com/datapoints/rideshare-industry-overview/#:~:text=Looking%20at%20the%20rideshare%20competitors,July%20and%20August%20of%202023) ## 1.2 Mobility Unit Economics We estimate that Uber’s existing take rate as of 2023 to be as high as 44%, assuming a cost breakeven scenario. An estimate of Uber’s unit economics at a ride level. Estimates were extracted from Uber’s Annual Financials in addition external analysts’ work. For a detailed description of these categories and the approach through which their values were obtained, refer to [Appendix A1](https://www.notion.so/Rideshare-Economics-Primer-Paul-s-Edit-Thad-Eugene-s-Work-Notes-e67b90ae714c47c69e216ed78887c00d?pvs=21): Uber’s take rate at 44% can be decomposed into 29.5% variable costs and 14.9% fixed costs. In relative terms, Uber is spending 66% of its revenue on variable costs and 34% on fixed costs. Based on our estimates, Uber’s take rate has been increasing from 34% to 44% from 2019 to 2023, which corroborates external findings. ### Variable Costs A large portion of Uber’s spend consists of variable costs for rideshare operations, more notedly, user acquisition (driver and rider). This spending is combined in the chart above under the “***Incentives & Advertising***” category which constitutes a significant 20.2% of a ride fare. Reducing or eliminating this category provides margin for drivers to be paid more and riders to pay less. The other variable costs (e.g., Support, Insurance, Credit Card (CC) Processing Fee) are relatively constant in their relative percentages to the ride fare. ### Fixed Costs At Uber’s scale, fixed costs are the cascading costs of a centralized organization. The following section describes these components: A substantial 5% is categorized as "Cost of Revenue (Overhead)". These are estimated from balances unaccounted for Remaining competitive in a centralized manner requires sustained growth of R&D (hiring and resources), which also increases General Administrative costs. Annual restructuring contributes to small costs. Together, these organizational costs take away 4.3% from each ride. Housing employees also incurs overhead costs like corporate office leases, real estate, utilities, and maintenance, categorized as Variable Lease. Additionally, there are asset-related costs: - Operating lease for Uber's vehicles (including light vehicles) - Amortization and depreciation of physical and intangible assets (e.g., computer equipment, furniture, land, software, etc.) - Asset impairments of mentioned assets This means that Uber spends 3.8% of ride fare on maintaining its assets. Uber's centralized approach to ridesharing across different regulations makes it vulnerable to legal liabilities. Therefore, part of their expenses goes to a "legal, tax, and regulatory reserve", which constituted 0.2% of gross bookings in 2023. This value fluctuates yearly, as reached as high as 1.4% in 2022. Since Uber controls the take rate algorithmically, any regulatory settlement costs can potentially be passed on to drivers.