Four questions [Adler frame]
Q1 — What is it about? Kevin Aluwi, co-founder and former CEO of Gojek, reflects on building Southeast Asia’s largest on-demand platform from a motorcycle taxi service into a publicly listed super app. The episode covers: why the super app strategy is weaker than its proponents claim; how Gojek survived through operational grit that competitors would not match; the role of brand in winning a capital-disadvantaged position; and the lessons of building outside established technology ecosystems.
Q2 — How is it argued? Aluwi argues primarily through case studies and counter-evidence drawn from his own operating experience. He does not offer abstract frameworks; he offers situations Gojek faced, the decision taken, and the result. The mobile top-up UXR finding and the motorcycle mafia story are doing most of the argumentative work. The super app critique is grounded in observed user behaviour (40% awareness of a top-six home-screen product), not theory.
Q3 — Is it true? The empirical claims are plausible but internally sourced. The 30–40% and 60–70% awareness figures for mobile top-up come from Gojek’s own UXR team and cannot be independently verified. The claim that Sequoia called Gojek’s growth the “craziest they’d ever seen” is anecdotal. The broader claim — that super app cross-sell economics are weaker than advertised — is directionally supported by public evidence: WeChat remains the rare exception, and most Western super app attempts (Uber, Grab, Revolut) have struggled to demonstrate the promised CAC savings. Aluwi’s framing matches the critical literature on platform bundling.
Q4 — What of it? The super app critique should cause product teams to question whether bundling is a genuine user benefit or a strategy-deck narrative. The more durable insight is operational: doing hard things that competitors avoid is itself a competitive position, and the barrier is will rather than capability. For founders outside Silicon Valley, the episode offers a realistic account of the compounding disadvantages (talent scarcity, capital scarcity, infrastructure gaps) and the specific adaptations (ops-heavy early stage, remote engineering early, don’t copy) that overcome them.
Glossary
Super app — A single mobile application offering a wide range of unrelated services (ride hailing, food delivery, payments, financial services, etc.) under one product surface. The thesis is that one trusted app with existing engagement reduces customer acquisition cost and enables cross-sell. Aluwi distinguishes this from a platform strategy (see below).
Cross-product synergy — The hypothesised benefit of bundling: lower customer acquisition cost because existing users discover new services, higher retention because the app becomes indispensable. Aluwi treats this as largely theoretical in practice.
Platform strategy — Making other businesses successful through shared infrastructure (think Stripe, Shopify, AWS). Aluwi contrasts this with the super app model: a platform adds services for others; a super app adds services for itself.
Mobile top-up — Prepaid airtime and data credit purchase. In Indonesia and much of Southeast Asia, the majority of mobile users are on prepaid plans and must purchase credits periodically. Gojek offered this as a product; 60–70% of customers did not know it existed despite it being on the home screen.
Gojek — Indonesian on-demand platform founded ~2010, originally motorcycle taxi (ojek) dispatch. At peak: ~30 services including ride hailing, food delivery, grocery, payments, and financial services. Merged with Tokopedia to form GoTo, which completed Indonesia’s largest IPO (c. $27–28B valuation).
GoTo — The merged entity of Gojek and Tokopedia (Indonesia’s largest e-commerce platform), listed on the Indonesian Stock Exchange.
The super app thesis and its limits
Aluwi opens his super app critique by naming the standard pitch: lower CAC through existing relationships, higher retention through indispensability, and easy cross-sell. He describes this as appealing on a strategy deck and weak in practice.
The mobile top-up evidence. Gojek offered prepaid mobile top-up — a product relevant to over 95% of Indonesian users, who are almost all on prepaid plans. It sat on one of six buttons on the home screen. UXR interviews found that only 30–40% of users knew it existed. Aluwi treats this as falsifying the naive super app assumption: engagement and visibility do not automatically transfer across services.
His explanation: users carry a mental model of what an app does. For Gojek, that model was “the driver.” Services that map to a motorcycle driver — ride, package delivery, food delivery, grocery delivery — were intuitive extensions. A driver could pick up food. A driver could deliver a parcel. Users could extend the mental model. Mobile top-up does not involve a driver; it breaks the cognitive link and must be sold almost from scratch.
The massage service illustration. When Gojek launched on-demand massages, users asked whether the driver would come into their home and perform the massage. The association was so strong it colonised a completely different service category.
The design constraint. Super apps converge on a grid or menu layout because no other UI can display 20+ unrelated services. This limits design quality and discoverability. Aluwi sees this as an unsolved problem rather than a solved one.
The implication for CAC. Because users do not automatically discover new services, the company must advertise and educate to drive adoption of each new service. This partially or fully eliminates the CAC advantage that is the central argument for the super app model.
Aluwi distinguishes his critique from a universal rejection: the model can work when services share a genuine unifying concept in the user’s mind. Gojek’s early expansion was successful precisely because all services mapped to the driver concept. The failure came from extending beyond that concept into services with no conceptual link.
Operational grit
Aluwi explicitly rejects the concept of moats as durable competitive advantages. His alternative: doing hard things competitors will not do is itself a position. The barrier is willingness, not capability.
Motorcycle taxi mafia. In Jakarta and other Indonesian cities, motorcycle taxi operations were organised and defended by local mafias who controlled specific areas through violence. When Gojek drivers began picking up passengers and orders in those areas, the mafias physically assaulted drivers: bricks, knives, machetes, mob violence. Aluwi’s response was to hire private security firms to patrol hotspots and extract drivers from dangerous situations. The company ran this operation for years until Gojek’s presence normalised motorcycle taxi services across the city. The signal to drivers: Gojek was not a platform that treated them as disposable contractors. This built driver loyalty that persisted when better-funded competitors arrived offering higher incentives.
Cash distribution centres. Digital payment infrastructure was embryonic in Indonesia when Gojek launched. Drivers needed to withdraw earnings in cash. Gojek built physical booths — a vault and cash, staffed — where drivers could present their driver ID and receive payment. This was essentially a manually operated mini-ATM network running at scale across a country of 17,000 islands. The company ran this until bank and ATM integrations became viable.
Stadium driver onboarding. Crystal Widjaja (mentioned but not detailed here) referenced a period when Gojek rented a stadium to onboard large numbers of drivers simultaneously, issuing phones and credentials in person because digital onboarding infrastructure did not yet exist.
Fraudulent driver apps. Third-party apps mimicked the Gojek driver interface and added features Gojek had not built (e.g., auto-accept orders) while also stealing driver credentials. Gojek lacked the engineering capacity to implement strong API security quickly. The response: identify the most popular features in the fraudulent apps and build them into the official app. Driver adoption of the fraudulent apps dropped. Aluwi describes this as a decision of necessity rather than principle — but it solved the problem faster than the security investment would have.
The pattern across all these examples: the response to a hard operational problem was to absorb it rather than abstract it away. Aluwi frames this as building capability and goodwill that compounds. A competitor with more capital but less willingness to operate in this way could not easily replicate these advantages.
Platform vs super app strategy
Aluwi draws a sharp distinction between two models that are often conflated:
- Super app: the company adds services to an existing app, capturing more of the user’s activity.
- Platform: the company provides infrastructure that lets others build services, capturing a share of their activity.
He does not explicitly argue one is superior, but his critique of super apps implicitly favours platform thinking: platforms do not require a cognitive link between services because the user never has to comprehend the platform’s full scope. A payment infrastructure platform serves food delivery, mortgage brokers, and ticketing without users needing to understand that relationship.
Gojek’s trajectory illustrates the tension. Early Gojek was closer to a platform for motorcycle drivers. As it added services beyond the driver concept — massages, movie tickets, loans — it drifted toward the super app model and encountered the awareness and CAC problems described above. The post-merger GoTo, combining Gojek’s on-demand services with Tokopedia’s e-commerce, is a further test of whether shared infrastructure (payments, logistics) can justify the combined surface in ways that pure super app bundling cannot.
The product strategy lesson Aluwi draws: if you add services, they must share a conceptual anchor that users already hold. If they do not, you are not building a super app — you are building several separate apps inside a shell.