Platform Scalability : Local Networks

It was a bright day during the winters of 2008. My co-passenger, a Kazakh boy of my age, and I connected over Bollywood ¹ and Shahrukh Khan while travelling from Berlin to Munich in a shared car. Like me, he was also backpacking in Europe, and we both loved using the German ride-sharing website Mitfahrgelegenheit.de ² – which translates to ‘drive-with-opportunity‘. He also introduced me to the French version of it – BlaBlaCar. In between our Bollywood banter, I could also sense an entrepreneur in him. He liked the idea of ride-sharing so much that he wanted to start something similar in his home country. Whereas, I firmly believed that despite a rise in the number of vehicles on the Indian roads and several other issues arising with them, carpooling for one-off or long-distance journeys would not succeed in India, at least then. In my perception, there were two main reasons. First, India’s internet penetration (% of the total population) was at a meagre 4.4% ³ in 2008. Second is the diverse socio-economic strata and the presence of multiple languages in India. Both of these challenges together could arise issues like reliability, safety, security and trust.

Few young entrepreneurs, who had different thoughts than mine, took some of these challenges head-ons and founded companies like Ola and TaxiForSure for urban passengers. Ola started as ‘Ola Trip‘ -a tour & travel operator in 2010 and soon pivoted to a ride-hailing platform. TaxiForSure began to operate as a taxi aggregator in 2011. In 2015, Ola acquired TaxiForSure to take on competitors.

Outside India, Ola now operates in UK, Australia & New Zealand. Its pre-COVID valuation was $5.5 billion.

San Francisco ➔ Bengaluru

Miles away, known for the ruthless & aggressive strategies, Travis Kalanick, the then CEO of Uber, also had high ambitions for the Indian subcontinent and considered India a “super-important” market. In August 2013, Uber started rolling out ‘Secret Ubers‘ ⁴ in Bengaluru to test the service. As is Uber’s tradition, it celebrated the launch by giving chauffeured rides to local celebrities. Soon, it expanded the stealth beta to other metro cities, and within a year, much cheaper options were made available to the public.

Although, after seven years, Uber claims ⁵ to command more than 50% of the ride-hailing market in India, the ride has not been less than bumpy for it in India. On top of the competition against homegrown, well-funded and aggressive rival – Ola , Uber faced many challenges that it didn’t face elsewhere.

Heterogeneity

India is very heterogeneous in terms of fixed and mobile internet usage by socioeconomic groups, age, gender, race, and India is very heterogeneous in terms of fixed and mobile internet usage by socio-economic groups, age, gender, race, and ethnicity. When Uber entered, internet penetration in India was around 15% ⁶. India had about 524 million mobile phone users (mainly concentrated towards the urban), but the penetration of smartphones as a percentage of mobile users was pegged at just 10%. The average internet speed was ~ 1.5 Mbps ⁷, one of the slowest in Asia-pacific and 117 internationally.

For a map-based app, technology was another bottleneck, not just because of the reliability of the internet, mainly because of the landscape of India. You cannot rely on GPS completely for ETAs and directions. The roads tend to be terrible; uneven, unlit, clogged with potholes, narrow, sharp turns, sudden stops, unscientific and unmarked speed-breakers, missing warning signs etc. Bengaluru, for example, is infamous for its one-ways and ever-shifting routes.

Another issue, which is more cultural – In India, people were not used to maps. They do not follow formal stricture; instead of cardinal directions, they would prefer to use a series of routes and familiar landmarks. Communication becomes utterly ineffective for people who don’t share a language. Over time, it became a norm to expect a direct phone call from your Uber or Ola driver confirming your destination after accepting the ride in the app. And if for any reason, like the traffic condition on that route or mode of payment, they don’t prefer to accept the ride, they would rather cancel the ride or ask you to cancel the ride.

Similarly, when it comes to payments, Uber’s just-get-out-of-the-car, aka cashless model in the US, was surely not going to work. Credit Card penetration was as low as two percent ⁸. So within two years, it had to roll out the cash option – Uber CASH (not to be confused with Uber Cash – digital payment within Uber Apps). Uber also adopted Paytm‘s digital wallet, allowing customers to deposit money into their accounts without plastic.

Similarly, when it comes to payments, Uber’s just-get-out-of-the-car, aka cashless model in the US, was surely not going to Many people (~ 80 %) preferred cash over app-based transactions until recently (post-demonetization & proliferation of digital payment apps). In one incident, before travelling to Guwahati, the capital city of Assam, in 2019, I was reminded multiple times to carry cash for my Uber ride payment. I didn’t want to take a chance, and instead of topping up my Uber Wallet, I queued for 30 minutes at the ATM outside the Airport. After booking the ride, the driver called me to confirm my destination, mode of payment and then instead of coming to the designated pickup point, he gave me directions to his parking location. Strangely, I was feeling lucky, but the experience had left me very curious. It was a long ride, and so I tried to get my answers from the Uber driver. When a rider pays digitally for their ride on Ola App, the driver is paid the next day through bank transfer and within a week in the case of Uber App. In both cases, the driver would need to visit a bank or ATM to withdraw cash which could take considerable time and might affect his weekly target of rides set by these companies. To avoid all this hassle, they prefer to take some money home daily. On the demand side, riders prefer cash to save the booking fee and additional charges levied by the bank.

Laws & Regulations

In addition to differences in culture, economy and demography, the expansion of digital platforms across geographies is In addition to differences in culture, economy and demography, the expansion of digital platforms across geographies is affected by regulations. Governments across the border differ in regulating foreign investments, competition, data privacy, net neutrality, intellectual properties etc.

Transport regulations in India are notorious for being complicated, allowing companies to interpret in ways that suit one’s interests. To complicate further, road transport in India comes under the domain of State administration. The central law — The Motor Vehicles Act regulates road transport vehicles, requires specific permits for transport vehicles, and stipulates various conditions and requirements for holding such licenses. The Act also grants state authorities the power to issue rules regulating taxis. State governments have established radio taxi systems that control traditional radio taxis’ operations in exercising this power. State government and taxi driver unions also regulate taxi fares. And then, there is the Information Technology Act, which provides the legal framework for IT companies in India.

To sum it up, there were no concrete central or state laws in 2013 to regulate digital aggregators. Uber used this for its benefit and ducked all the existing regulations. RBI (India’s central bank) also found Uber in violation because of using ways to bypass Indian laws to conduct transactions overseas. Uber recruited drivers for its technology platform without any background checks and, in many cases, without permits to carry local fares. It soon faced its first legal ban in India in Delhi on December 8, 2014, after a woman accused an Uber driver of raping her. The investigative agencies took hours to locate Uber executives as apparently the company was operating from few hotel rooms. It took additional 5 hours to get the details of the accused driver as the servers were in the US.

Still recovering from the 2012 Nirbhaya case, these incidents proved to be a wake-up call for India. As a result, Delhi Government modified its existing radio Taxi licensing rules to allow app-based taxis aggregator to be eligible for a radio taxi license. It hence will have to abide by all the relevant statutes, including the central Motor Vehicles Act and Information Technology Act. The app-based taxis aggregator like Uber were now :

• Mandated to introduce safety measures including SOS, live vehicle tracking, dedicated control centre and call-centre for distress signals
• Responsible for the quality of drivers, police verifications, their conduct with passengers
• Required to have a registered office in Delhi with details submitted to the Transport Department.

Soon other states also tightened the regulatory noose around ride-hailing apps’ neck, and presumably the reasons for the slow expansion of Uber. As of 2021, Uber operates in 89 cities in India, whereas Ola operates in more than 250 Cities. In 2020, Central Government issued new guidelines to the app-based ride-hailing companies setting the rules for commission & surge pricing.

Uber is not new to running into legal tangles because of its employees and drivers conduct. Parallelly, it keeps getting into difficulties with the regulatory authorities. Besides the US & India, Uber has been contending regulatory battles in many countries — The Netherlands, France, Spain, Germany, UK, China, Thailand, Vietnam, South Korea, to name a few.

While this may not be entirely true that regulations imply a slower growth for foreign platforms, the differences by country indeed threaten their success. In contrast, a local platform may benefit and prosper. To stay out of trouble and to continue to thrive, such platforms needs to adapt to local peculiarities.

Network Effects

For a local marketplace, like Uber, co-location or geographical proximity of the supply side and the demand side is essential. Adding a supply unit (driver) makes the product more valuable to the demand side ( riders) within a small geographic radius and vice versa. It is of less or no value to the riders if the drivers are beyond this radius. Similarly, for the drivers, it is entirely immaterial if there are more riders in another city. A network effect, where the value of the network to its users increases with the number of users in proximity, is known as ‘Local Network Effect‘.

In the case of Uber, local networks effects come into play. Therefore, whenever it expanded to other cities (in the US or any other part of the world), it first had to attract a new fleet of drivers, and once a critical mass of drivers is on board, the second cycle starts to attract the riders. The acquisition of drivers & riders requires additional expenses and investments; the only advantage is that they didn’t have to make significant investments in certain reusable assets such as their IP, i.e. match-making algorithm. To attract & retain drivers, Uber incentivized them with monetary rewards in addition to per-ride compensation. And to attract riders, Uber advertised heavily, threw discounts and subsidized the cost of a ride.

Uber is primarily dependent upon the untapped labor and capital to compete with private car ownership in the US. In contrast, in India, it had to leapfrog car ownerships and fulfil the supply by recruiting and training drivers (including and not limited to driving, smartphone usage and customer service). Most drivers don’t own cars; they must buy or drive someone else’s vehicle. To get more cars and drivers on the platform, Uber had to set up leasing deals with banks, finance companies and car manufacturers to get special terms for the drivers in need of vehicles. Consequently, a significant part of the driver’s earnings goes towards repaying the loan. The other set of complexity in acquiring drivers arises due to local competitors in the market. In India, established local players like Ola and taxi aggregators like Meru Cabs, EasyTaxi, made hiring trained drivers more complicated, expensive, and daunting. Either way, such an arrangement was bound to result in disgruntled drivers, particularly when their incentives and subsidies were gone.

Uber believes that its scale gives it a ‘liquidity network effect‘, their most significant asset. In their S-1 , they emphasized that –

“the foundation of our platform is our massive network.“

They also claim that the network effect will lead to a long term competitive advantage.

“Our strategy is to create the largest network in each market so that we can have the greatest liquidity network effect, which we believe leads to a margin advantage.“

As per the diagram above, included in Uber’s S-1, they claim to have 2-sided marketplace network effects, but in reality, Uber’s network effects are asymptotic. Asymptotic networks aren’t strong, as the benefits of increasing the supply beyond an optimal wait time have diminishing returns.

Two other factors weaken the ride-hailing company’s competitive advantage in an existing market:

• Low Switching Costs:  It is no mystery that the participant on both sides (drivers and riders ) of a ride-sharing platform multi home. It doesn’t cost a dime to install an additional app, and then it’s just a matter of few swipes to switch from one app to another.

• Low barriers to entry: Due to local network effects and absence of scale effects, new entrants can make inroads and gain a critical mass of supply-side by focusing on single geography. Uber’s global presence and millions of drivers is immaterial in such a case.

Ride-sharing apps like Uber know these weaknesses very well, and that is why they have been building defensibilities around the weak but extensive network. Uber has consistently added new products as “nodes” to its network to reinforce its weak network effects. Some of the notable ones are:

• Uber Eats
• Uber Freight
• Uber Commute
• Uber for Business
• Uber Direct
• Uber Connect

Uber’s Numbers

Uber has come a long way. It has diversified into other adjacent markets by leveraging its network of drivers and logistic capabilities – Food & grocery delivery (Uber Eats), P2P delivery (Uber Connect, after shutting down Uber RUSH ), on-demand goods delivery (Uber Direct) , a completely new shipping platform that connects trucks drivers and shippers (Uber Freight) and a self-driving unit (Uber ATG, which was sold to Aurora in 2020). Some of these activities, like Uber Eats, might define Uber’s future path, but currently, they are in a nascent stage and not profitable. Ride-sharing remains the core of Uber, and the struggle to keep the costs low continues. The main cost drivers for the core platform are constant expansion to other markets, high cost of customer & partner acquisition and high churn rate of drivers.

Worldwide, Uber might be a market leader in ride-hailing, but its business model is a loss-making one, and that’s not a secret. Uber lost around $2 billion in 2015 on gross revenues of $ 1.3 billion, and in 2016, the losses grew by 175%. Until 2017 it lost more than $11 billion cumulatively. 2018 was the only year it was profitable on paper, probably due to its exits from Russia, China, & Southeast Asia. The same year it also shut down Uber Rush. Uber went public in 2019; the reported losses that year were $8.5 billion. During the pandemic, the YoY net losses came down by 14%.

Compared to 123% in 2017, Uber’s Core Platform Adjusted Net Revenue (sans driver incentives) grew by just 39.4% in 2018 and declined YoY since then. Uber’s Gross Bookings (money spent by users on the platform) increased un-impressively, but its Take Rate – the percentage of Gross Bookings it captures as Core Platform Adjusted Net Revenue, has been in steady decline from 2017. 2017 was a disastrous year from the PR and market perspective. One thing that came out very clearly was that Uber couldn’t achieve profitability by reducing the driver incentives. As the world recovers from the pandemic, a significant increase in the investments in driver incentives to improve driver availability is required. The recent rise in fares by ride-sharing apps indicates that, but the question is, how much of this increased fare the drivers will get?

Well, the story isn’t that gloomy for Uber as I might have portrayed it here. It has seen success in markets like the Middle East, where it acquired the ride-hailing company Careem. Compared to other US tech companies like eBay & Google, Uber did a few things right in China, but still, it was challenging to unseat a fiercest local rival, Didi. Uber might have surrendered in many battles, but its stake in its competitors might be worth $18 billion or more, out of which more than $5 billion is from Grab‘s recent $40 billion valuations. Most importantly, a presence and large user base worldwide has unlocked scale advantages similar to a global marketplace for the company.

It’s evident from its expansion efforts in India, Southeast Asia, China & Russia that Travis Kalanick vision of ‘remaking transport‘ needed more than copy-pasting. A proven technology solution that works in the first world isn’t enough to operate in another economy esp. emerging economy. The product needs to be thoughtfully localized and personalized. Uber’s unique product features like: In-app emergency aka panic button, vehicle tracking, arrears handling, driver inbound support, ‘Dial a Uber‘, UberDost, Uber CASH, Uber Rentals and UberGO originated in India from the local needs.

Local networks like Uber aren’t bad, but they are geographically constrained. Such constraints limit profitability because of the amount of funding they require for expansion. Therefore you need to be thoughtful in your geographical expansion. Uber’s rivals – Bolt & Lyft, also remains unprofitable but are focused on operations and costs.

Here are some key takeaways for the Local networks:

• Reach critical mass on both sides as fast as possible in strategically important localities
• Focus on expanding to enough geographies which are similar in terms of culture and regulations.
• Prioritize to geographies that are high density and less competitive.
• Invest more on the constrained side first to attain a dominant position in the market.
• Build mechanism to retain the constrained side and prevent multi-tenanting

---

1. Fun fact – Bollywood movies are so popular in Kazhakstan that they used to have a dedicated 24 hour channel for Indian movies. Bollywood Actor – Mithun’s dance on Jimmy Jimmy is more popular in Khzhakstan than in India.
2. Mitfahrgelegenheit.de founded in 2001, was rebranded as CarPooling.com in 2011. In 2015, it was acquired by BlaBlaCar (founded in France 2006). BlaBlaCar entered in Indian market in 2015.
3. World Bank Data
4. Uber Blog
5. Uber Investor Presentation – Feb 2020
6. World Bank Data
7. The State of Internet -By Akamai Research
8. E-commerce Payments Trends: India – By JP Morgan