On a rainy Sunday in Jakarta, a dispatcher stood outside a shopping mall in the eastern part of the city, asking the approaching customers how many bike taxis they would need. “Four!” he shouted, turning to the group of drivers waiting nearby. Four motorbike-taxi drivers stamped out their cigarettes and quickly gulped their coffees to catch their next fare. Each passenger told their driver the intended destination, swung a leg over the backseat, and took off, purchases in hand.
These passengers were not seeking rides at a traditional bike-taxi stand. They were at GrabNow shelters, mobility platform Grab’s latest “innovation” in Indonesia. The rules here are simple: drivers turn their apps off so they can’t be booked online, write their names on a Grab-provided whiteboard, and wait to be summoned by a Grab-hired dispatcher. Once it’s their turn, they erase their name from the whiteboard and scan their customer’s QR code or enter in a numeric code from the customer’s Grab app.
The idea of a bike-taxi station is, of course, nothing new. Well before apps like the Singapore-based Grab came to Jakarta in 2015, hundreds of “offline” bike-taxi stations served customers across the city. After a few years of disrupting this market, Grab began embracing the practices of street hailing and offline queuing — the very systems it had once purported to replace.
Startups like Grab, which position themselves as “homegrown” disruptors, are often just replicating Silicon Valley–inspired algorithmic solutions in a new context. More often than not, their dream of tech-enabled, frictionless consumption falls apart when solutions move from the boardroom to city streets.
Part of Grab’s pivot likely came about because ride-hailing technology in Jakarta didn’t work as well as it was expected to. In many areas of the city, gridlocked traffic makes online matching much less efficient than simply hailing a bike. In the time a matched motorbike 500 meters away navigates traffic and responds to an order, a customer can hop on any one of multiple available motorbike taxis waiting by the side of the road.
The efficiency of ride-hailing platforms relies on drivers who are open to being matched anywhere actively searching for orders in different parts of the city. In Jakarta, however, by 2016, spontaneous stations started popping up only for app-based drivers. Dozens of Grab or Gojek drivers would congregate on the sides of roads, waiting for the platform to match them to an order. This practice was a reversion to the old bike-taxi model, in which every driver had a station he or she belonged to. As drivers became more attached to stations, mobility platform companies had trouble incentivizing drivers to move away from these anchored spots.
Grab positioned these changes as a triumph of innovative problem solving, not a failure of its model. When he launched GrabNow in 2017, CEO and co-founder Anthony Tan celebrated the company’s successful innovation during a press conference in Singapore: “Previously, you had to book, wait, look at it, and spend a few minutes. Today in Jakarta, you see, you go immediately.” The irony was that Jakartans had found rides “immediately” in the traditional bike-taxi market, before Grab’s entry.
Some might say that GrabNow is a successful example of localization: a platform adapting to its market. The assumption that ride-hailing technology would operate in Jakarta as it does in any other city in the world, however, is representative of the tech industry’s misplaced faith in one-size-fits-all solutions. As the pandemic continues to give rise to a menagerie of startups offering e-commerce and delivery solutions, we must remember that technology can never succeed in isolation.
This dynamic is not just limited to Indonesia. Platforms constantly take on different personas as they move to different markets, adjusting for regulatory problems or algorithmic failures that were unanticipated (or dismissed) by those in charge. In Pakistan, the people wealthy enough to own cars are unwilling to drive them for profit. As a result, ride-hailing platforms became a conduit for brokers who would hire drivers to work 12-hour shifts for minimum wage. In South Africa, violent clashes between taxis and Uber drivers forced Uber to deploy armed private security in demand hotspots. The story of technology, many forget, is not just about the app or the algorithm, but also the context.
Even local startups cannot truly localize if they model themselves on Silicon Valley successes. An ever-expanding list of startups in emerging markets –– Getir, Rappi, Bykea –– now offer mobility, delivery, and e-commerce options. Much like their Silicon Valley counterparts, these startups believe the greatest problem to solve in a market is logistics: matching consumers to services. Depending on the context, a car might be transformed into a bike or a van, the app might operate in Urdu or Turkish, but the underlying assumption remains the same: that existing urban systems are less efficient than their tech-based counterparts.
When designers solve for only certain problems, it often falls on ignored users to repurpose those technologies. Researchers such as Payal Arora, Ramesh Srinivasan, and Nanjala Nyabola have shown how users in non-Western markets adapt digital technologies for local problems. WhatsApp was conceived as a social chat platform, but street vendors in India use it as an e-commerce app to update customers on their wares. Meanwhile, government departments in Pakistan use WhatsApp to monitor taxes, and doctors in Brazil see it as an e-health platform. In Rwanda and Sri Lanka, phone owners save money by communicating via missed calls. Instead of paying for minutes, people will hang up after a single ring to deliver a predetermined message: pick me up, call me back, I am safe.
To truly localize tech, as many startups in emerging markets claim to do, we cannot simply change the language of the operating software and call it a day. Entrepreneurs must instead ask if their solution is necessary or appropriate. Designers must question whose needs they are centering. VCs must ask if the problem being solved is even a problem in the first place.
Silicon Valley’s brand of thinking –– fixated on technology as a solution, bolstered by exploitative labor practices, marked by a narrow definition of innovation –– should be a cautionary tale, not a model to follow.