Careem's first mover disadvantage in Pakistan
A case of failure from failing to self-disrupt
Careem, along with Uber, pioneered the ride-hailing market in Pakistan.
From 2016-2020, both companies spent millions of dollars to establish the ride-hailing industry in Pakistan. To support the passenger-side, both companies offered subsidized pricing and gave away extensive promotions to convince people to travel in cars with strangers. On the driver side, they ran extensive sales operations and offered lucrative incentives to convince car owners and drivers to operate their cars on the platform.
At the time of Careem's acquisition by Uber in Q1 ’2019, collectively both companies would have had ~90% market share of Pakistan's ride-hailing market.
Yet, by mid-2025, InDrive and Yango reduced this market-share to single-digit percentage points, prompting Uber, which had already folded its own Pakistan operations in Careem’s favor and fully owned Careem Rides, to end Careem's ride-hailing operations in Pakistan.
My thesis is that this is a classic case of business model disruption. Careem/Uber failed to keep up when the ride-hailing market in Pakistan evolved from being passenger-led to driver-led.
Let’s examine how.
Establishing multi-sided marketplaces is a challenging endeavor. A new entrant has to constantly prioritize and re-prioritize their limited capital and management bandwidth between growing one-side of the marketplace versus the other.
Prior to Careem entering Pakistan, many major cities in Pakistan did not even have a formal taxi market. Women generally avoided using public transport for cultural and safety reasons. To introduce the ride-hailing concept, Careem prioritized the passenger-side of the marketplace when it first entered in 2016. To establish its brand and convince people, especially women, to travel with strangers, Careem launched with subsidized pricing, extensive promotional discounts, vehicle inspections, driver background checks and 24/7 call-center access for support.
In order to have a vehicle available to any prospective passenger within a 5 minute ETA anywhere in the city, a large number of drivers always needed to be present. To convince vehicle owners and drivers to work with the platform, Careem enticed them with guaranteed earnings, large bonuses and payments for hours being available even if they didn’t get a booking.
All of the above required a tremendous amount of investment and in the first couple of years, Careem was burning hundreds of thousands of dollars a month in the Pakistan market. This period from 2016 to early-2019 also coincided with Uber, the global well-funded David, and Careem, the scrappy regional Goliath, engaging in a ruthless competitive battle to drive out the other company and win a monopoly position in the industry.
The plan was that gradually as a critical mass of passengers and drivers started relying on the ride-hailing platform as their primary sources of transportation-services and income respectively, the platform would start raising prices and reducing promotional discounts on the passenger-side, reduce subsidies on the driver-side and carve out its 20-25% commission between the price that the passenger pays and the revenue that the driver gets.
In April 2019, Uber announced its acquisition of Careem which ended their death match. A couple years later, Uber started closing the Pakistan operations under its own brand in favor of Careem. By 2020-2021, after the initial Covid shock, Careem had been fairly successful in ending passenger and driver subsidies, started earning its operational margins and had made its Pakistan operations cash-flow positive.
The problem with subsidies delivered via taxpayer funds or venture capital dollars is that they distort market signals and bring individuals to the marketplace who may not be able to afford the true economic prices. Though for a new entrant looking to change people’s behavior and drive adoption for a nascent industry, subsidies can be useful in helping people overcome their status-quo bias and try out a new offering. Subsides are like a drug, an artificial stimulant. When they are eventually revoked they leave behind an unhappy population suffering from withdrawal symptoms. Whether that’s in the form of a young professional who took a job commuting far from home due to the availability of cheap, ride-hailing services and is now facing a loss of take-home earnings when his commute costs rise; or in the form of a driver that leased a car with the expectation to use the earnings from the ride-hailing platform to pay off the car-loan and is now facing difficulties when the ride-hailing platform stops paying out what it used to.
Thus by 2021, Uber/Careem had invested millions of dollars to establish the ride-hailing industry in Pakistan but had cornered itself into the precarious position of having a large portion of their customer and supplier bases feeling “things aren’t as good as they used to be”. They had created ripe-conditions for a new entrant / second-mover like InDrive to come in and take the cake.
Careem and Uber as the marketplace operators provided value to the marketplace along the following vectors, ranked in order of importance:
Ensuring availability of a vehicle within a 2-5 minutes ETA.
Price determination, including the controversial surge / peak charges that were designed to incentivize supply during periods of high-demand.
Payment processing, Accepting card payments from riders and then paying out to drivers.
Customer support operations, including a 24/7 priority access for active passengers.
Ensuring an acceptable quality of drivers and vehicles on the platform.
For these services, the platform would charge a 20-25% commission or take-rate between what the passenger pays and what the driver gets. This managed-marketplace strategy is biased towards the passenger.
InDrive made the strategic bet that it could cut out a lot of these services and charge a much lower commission / take-rate.
In comparison, InDrive’s value proposition in the marketplace, is a lot more laissez-faire and as follows:
InDrive would attempt to ensure strong liquidity on the marketplace by having a large number of drivers active on the platform at any time, but it would allow passengers to make a trade-off decision between price and availability.
The InDrive platform will only suggest a price to the passenger, as opposed to enforcing it in the case of Uber/Careem. The passenger will make an offer and in-return receive bids from nearby drivers, and then choose the most suitable bid based on price, driver history, rating and ETA.
No card-payment processing option. Processing card-payments at costs of 1.5 ~ 2% of gross-transaction value would end up costing platforms 6% - 8% of their take-rates. With InDrive, passengers would have to pay the driver via cash or a direct transfer to the driver’s mobile wallet.
No call-centre support to deal with emergency situations, only an asynchronous help-desk for support.
Lesser policing of the marketplace for driver and vehicle quality. Easy, onboarding process for drivers.
For this free-market strategy, the InDrive platform would charge a ~13% commission and is generally biased towards drivers.
InDrive’s feature to give drivers agency in pricing by allowing them to place their bids proved to be a hit. The lower platform commission rate meant higher earnings for the driver. Though InDrive did expend some capital initially to offer enticing bonuses to drivers to convince them to switch from Uber & Careem, it faced no troubles in finding product-market fit in an established industry. 2021 and 2022 saw a mass-exodus of drivers away from Careem as their primary platforms to InDrive becoming the primary platform.
The drivers also convinced passengers to switch over to the inDrive platform via word-of-mouth. Drivers would take bookings on Careem and convince the Careem passengers to try out the InDrive platform.
As soon as a critical mass of drivers left the Careem platform, the availability and ETAs on Careem sharply deteriorated which forced passengers to switch over to InDrive. InDrive’s virtuous cycle of more drivers -> high availability and low ETAs -> more passenger requests -> more driver utilization & earnings -> more drivers became Careem’s vicious cycle of less drivers -> low availability and high ETAs -> less passenger requests -> less driver utilization & earnings -> less drivers.
InDrive’s platform being relatively inferior in value offered to passengers did not hold back InDrive’s growth. Passengers followed drivers to move over to the InDrive platform despite the lack of card payments being annoying for discerning users, lack of real-time support in special situations being frustrating and the ride booking mechanism being more complicated due to the offer -> bidding -> bid-acceptance process. The ride-hailing industry in Pakistan transitioned from being passenger-led to driver-led. Careem failed to stay ahead of this trend and ultimately got left behind in the industry it had a massive role establishing in Pakistan.
Careem tried to fight-back but its response was too little and too late. It introduced a similar flexi variant where prices would be determined via the offer-bid process and dropped commission rates to 15% in-order to try to keep up with InDrive. Yet it failed to stop its market share bleeding.
By the summer of 2025, InDrive had a roughly 60-65% market share in Pakistan, followed by Yango with ~30% whilst the industry pioneer Careem had dropped to less than 5%. Ultimately, Uber which owns Careem’s ride-hailing operations to announce Careem’s exit from Pakistan’s ride-hailing industry by July 2025.
PS This ties in with a broader concept I have been thinking about lately. Whilst any business comprises many functions from Sales, Product Development, Engineering / R&D, Design, Marketing, Support, Growth, Finance etc. There’s often only a single source of “alpha”.
A company that doubles-down and obsessively focuses on its source of alpha in the market would generally win in the long-term over companies that try to be world-class in all functional area and don’t allocate disproportionate resources to its source of alpha. For a more relatable analogy, if you’re exceptionally talented in languages, writing and creative expression, your odds of winning the Pulitzer Prize are lower if your limited time and cognitive bandwidth is equally split between writing and mathematics, where you’re average at best.
For the ride-hailing market in Pakistan, initially we saw the source of alpha coming from winning over the passenger-side of the marketplace. Aggregate demand on your platform and supply would follow. As the industry matured, the source of alpha shifted towards the becoming the primary platform for drivers. Have supply on your platform that leads to high availability and low ETAs and demand would follow.
I plan to elaborate more on this concept in a future post. Please subscribe to stay tuned.
Interesting read- much needed penning down of introspection for growth of market and learning.
I agree that subsidised pricing for riders set their expectations wrong- the pricing was aiming to cater to mass market (which we know is still too high for masses) and increasing pricing to cater to a defined target audience was a scary topic i.e. demand will go away.
In parallel, the drivers after the incentives (temp financials) were removed (after Uber IPO because population and ease of onboarding, due to loose regs in Pak meant you can show high number of drivers by pumping drivers in Pak to the overall global numbers) were left with unrealistic expectations and ride-hailing players were not able to solve for "what should partner economics look like". This meant a fire-fight loop of high early churn, continuous acquisition, training, fraud, churn and back to referrals.