Rocket Internet Gave It Eighteen Months

The Machine and Its Purpose

Rocket Internet's method was not subtle. The Berlin-based venture builder had spent years copying proven consumer-internet models — typically American or European originals — and deploying them simultaneously in markets where the original had not yet arrived. Speed was the product. If you could reach a market six months before a better-funded competitor noticed it existed, you could claim the liquidity advantage: enough matched trips to make the platform useful, enough users to make the switching cost real. Rocket's approach to timing is documented across its publicly filed prospectuses from its 2014 Frankfurt IPO, which describe the strategy in terms of "proven business models" and geographic arbitrage rather than invention.

Tripda launched in 2014, with BlaBlaCar already the dominant intercity carpooling platform in Western Europe and growing fast toward Latin America. The template was visible: a marketplace connecting drivers with empty seats to passengers willing to share fuel costs, mediated by a reputation system, priced at a legal cost-share rather than a commercial fare. What Rocket needed to know was whether that template would perform in markets like Brazil, Mexico, Indonesia and across Southeast Asia — markets with different infrastructure, different car-ownership patterns and different baseline trust in strangers.

The Arithmetic That Had to Work

Every intercity carpooling marketplace rests on the same fundamental arithmetic. A driver makes a journey anyway — the marginal cost of carrying an extra passenger is close to zero. The passenger pays a fraction of what the train or plane would cost. The platform takes a service fee, typically in the range of ten to fifteen percent of the transaction. The model only becomes a business when the occupancy rate — the share of empty seats that are actually filled — reaches a level where the platform's take covers its cost of matching.

That cost is not small in a new market. Average vehicle occupancy on long-distance road journeys in most countries sits close to one and a half people per car, which means that on a four-seat vehicle, roughly two and a half seats are making the trip empty on every journey. The pool of potentially shareable kilometres is enormous. The problem is aggregating it: a driver going from São Paulo to Campinas on a Wednesday afternoon does not help a passenger needing to travel that corridor on Thursday morning. The marketplace needs simultaneous density on the same corridors at the same times — and density takes time, marketing spend and user trust to build.

Rocket's bet was that time could be purchased. A large launch budget, deployed simultaneously across many markets, was designed to buy density quickly rather than let it grow organically. The theory was that a well-funded simultaneous launch would compress the trust-building and habit-formation process into a shorter window, after which the platform would reach self-sustaining liquidity.

Trust on an Accelerated Schedule

The trust problem was not incidental. The structural mechanism that made intercity carpooling function in Europe — where BlaBlaCar had been running since 2006 under its earlier French incarnation as Covoiturage.fr — was a layered reputation system: verified profiles, mutual ratings after each trip, and the gradual accumulation of a digital record that made strangers legible to each other. That system only works when there are enough previous trips for ratings to mean something, which is precisely what a new market lacks.

Tripda deployed the same toolset: profile verification, post-trip ratings, in-app messaging. The architecture was correct. The question was whether the reputation system could bootstrap fast enough in markets where the concept of sharing a car with an unknown person was not already culturally established. In Brazil, for instance, informal ride-sharing arrangements existed but were not standardised or recorded. The gap between existing informal practice and a rated, verified marketplace was both the opportunity and the friction.

Brazil was one of Tripda's larger bets, and the country's geography gave the model real structural support: distances between cities are long, intercity bus networks are extensive but not cheap, and car ownership concentrated in urban middle classes who regularly drove intercity routes. The detour cost — the extra distance and time a driver accepts to pick up a passenger — was manageable on long corridors where the shared leg was the dominant part of the journey. The economics were plausible.

What Eighteen Months Measures

Rocket Internet is explicit about the logic of its deployment timeline. The question a simultaneous multi-market launch is designed to answer is not "can this work?" — that question was answered by the European original. The question is "how fast does this market reach the liquidity threshold, and at what acquisition cost per trip?" A market that reaches critical mass in twelve months at an acceptable cost is retained and grown. A market that is still burning acquisition subsidy at eighteen months without approaching self-sustaining density is a signal, not an accident.

By 2015, BlaBlaCar had moved aggressively into Latin America and was acquiring local competitors rather than building from scratch. BlaBlaCar entered Brazil directly in 2015, buying the then-dominant local platform and consolidating the market in a single transaction. The dynamic Tripda's launch had been designed to pre-empt — the arrival of a better-capitalised competitor with a proven product — arrived faster than the window Rocket had projected.

Germany, where Tripda never operated, had already shown how a separate structural shift could undercut the model's pricing advantage. The 2013 deregulation of the long-distance coach market had filled the cheap-seat gap that intercity carpooling occupied, with FlixBus and others offering scheduled, insured, comfortable travel at prices that competed directly with a cost-share seat in a stranger's car. Building carpooling density in a market where the coach has just become the functional substitute is a particularly difficult position.

Tripda closed in 2016. The closure followed the pattern Rocket applies when a market test produces a clear answer: the platform stopped accepting new bookings, users were notified, and open bookings were managed through a defined wind-down window. The recorded outcome was not that intercity carpooling was wrong as a model — BlaBlaCar's continued growth made that argument difficult — but that the specific combination of markets, timing and competitive dynamics did not produce self-sustaining liquidity within the parameters Rocket had allocated.

What the Tripda experiment actually produced was a set of market readings: which corridors were structurally viable, what acquisition cost per active user looked like in an emerging economy, how quickly reputation systems bootstrap without a pre-existing culture of rated sharing. Those readings are the product a launch machine is actually built to generate. The closed platform is the invoice; the data is what was purchased.

Tripda launched across a dozen-odd countries in 2014 and closed in 2016 — a deliberate machine for testing a model quickly in many markets at once.