Markus Villig on Building Bolt, Marketplace Growth and Self-Driving

Harry Stebbings asks Bolt founder and CEO Markus Villig how he started the company as a 19-year-old in Estonia. Villig says he had wanted to be a technology entrepreneur since childhood, sold things at school and learned to code. His parents encouraged him, having grown up when entrepreneurship was prohibited under Soviet rule.

After an unsuccessful school-software experiment, Villig studied industries and chose transportation. He anticipated a shift toward on-demand services, electric vehicles, micromobility and eventually self-driving. He saw transportation as a place undergoing generational change, though the future he imagined was not yet established.

He searched online for startup advice and surveyed potential customers before coding. Hundreds described a poor Tallinn taxi experience: calls went unanswered, cars did not arrive and card payments were uncommon. Recruiting drivers was harder: he says about 90% initially told him to leave their cars, but after months he persuaded around 50 to share their details.

His pitch was that drivers sitting idle could earn additional trips for a small commission. Villig then built parts of the rider app, driver app and backend, but realized he needed an experienced engineer. His older brother Martin, an early Skype employee, helped search; Villig says at least 30 potential co-founders declined.

A candidate found through an Estonian online forum met Villig and returned days later with a working prototype. That engineer, Oliver, became the technical co-founder and remained with the company 11 years later. Villig credits luck as well as persistence, and recommends searching beyond one’s immediate network.

Ride-hailing presented a difficult marketplace problem: drivers expect trips quickly and riders expect a car within minutes. Villig’s parents provided the $5,000 they had saved for his university expenses, on the condition he would cover his own rent if he spent it. The money funded early development and simple marketing materials.

Free publicity brought customers from the first day, but too few drivers were online and many riders had a poor experience. Villig went back to taxi stands, set up driver accounts personally and recruited help from his brother and mother. He recalls about five trips on day one, 100 daily trips after a few months, and eventually thousands.

He says consumer demand was visible early, while supply remained the company’s persistent constraint. Bolt’s internal refrain became “add supply,” later encompassing couriers and restaurants as well as drivers. In hindsight, he thinks a small angel round sooner could have accelerated the first year.

Villig attributes Bolt’s frugal start to a team focused on reaching sustainability with very little money. Salaries were modest, he says, but equity helped attract committed employees. The cash shortage also encouraged detailed measurement of spending and returns; his comparison with competitors is his own account.

He says Bolt reached roughly €10 million in annualized revenue after raising about €1 million in seed funding. The early round came mainly from small Estonian funds and former Skype employees, at a valuation he recalls around €9 million. These are his recollections and annualized measures, not audited revenue figures.

The first expansion attempt launched ten countries in parallel and nearly bankrupted the company within six months, Villig says. Bolt shut those markets, let people go and returned to roughly 15 employees, with Estonia as the functioning base. It then learned to expand sequentially into nearby countries.

New ride-hailing markets generally begin with negative unit economics, Villig says, because both drivers and customers need incentives before there is enough liquidity. He estimates that around 25% market share is often needed for the network to become self-sustaining. That threshold is his rule of thumb, not a universal guarantee of profitability.

Martin launched Latvia from a rented apartment that doubled as an office, meeting taxi companies and drivers while distributing cards and leaflets. Villig says word of mouth accounts for about 80% of volume in a mature market, while paid channels make up the remainder; at launch he relies on media coverage and online advertising to seed the network.

In some early markets Bolt lacked money to subsidize drivers, so it first offered taxi companies fleet management and dispatch software. Once their drivers were on the platform, Bolt could build consumer demand over the following months. Villig describes this as getting users through a useful tool and retaining them through a network.

The biggest surprise was Africa. Villig says several European countries were each doing roughly €10 million in annual gross bookings, from which Bolt took about a 15% commission, or roughly €1.5 million in annualized revenue per country. Gross bookings represent customer spending, not Bolt’s revenue or profit.

The team ranked 200 cities using simple criteria including population, regulation, drivers and car ownership. African cities ranked highly, with Johannesburg first. Before hiring local teams, Bolt ran small online ads in prospective cities to measure sign-ups and acquisition costs, even advertising that the service was live before launching it.

In Johannesburg, Bolt recruited a student remotely to find an office and contact and train drivers. Villig says that within six months the city accounted for more than half of the company’s gross bookings, despite Bolt operating in roughly 15 European cities. He attributes the result to high trip volume and fares closer to Eastern European levels than expected.

Stebbings questions using gross bookings as the principal measure, because it does not directly show user satisfaction. Villig says retention, trip frequency and referrals ultimately appear in gross bookings, though specialist teams still track acquisition and retention. This is his management preference; gross bookings alone do not establish profitability or customer experience.

Bolt repeated the remote launch exercise, testing demand with ads and expanding into around 15 countries beyond its early European base. Villig says the company entered many markets after a competitor and had to overcome the incumbent’s larger driver network and shorter pickup times.

His explanation for catching up is localization and cost: for example, he says a competitor accepted only credit cards in Kenya, leaving many potential customers out. He claims Bolt’s smaller commission enabled lower rider prices and better driver pay. These are company claims and illustrative comparisons, not independently verified market-wide results.

Stebbings relays Uber’s question about whether being second or third is viable. Villig says Bolt operated in about 50 countries and was the most popular platform in more than 20, while expecting to gain share elsewhere. He argues a lean number-two operator can be profitable and speculates that Lyft would struggle independently without lowering costs; those competitive assessments and forecasts are his own.

Behind the expansion was a small Estonian team trying to solve localization, payments, invoicing, hiring and marketing remotely. Villig says one neglected area was public policy: in markets where Bolt helped define the category, regulators needed input earlier. He wishes the company had built that capability one or two years sooner.

Speed matters only when execution quality is high, he argues; moving fast by cutting corners can create momentum without progress. Looking back, he would have chosen expansion markets more deliberately instead of attempting the first ten simultaneously.

Despite rapid growth, Villig says Bolt often had only one or two months of cash and struggled to pay the next month’s salaries. He recalls meeting dozens, perhaps a hundred, European investors who passed, frequently because they believed ride-hailing was winner-take-all. His account of revenue and growth at the time is self-reported.

To keep going, he approached local businesses outside the venture industry. An Estonian real-estate company invested $500,000 at an estimated $15 million valuation, and other local investors added smaller amounts. Villig says the investment later returned more than 100 times; the interview does not establish the precise realized proceeds or timing.

Even at roughly $25 million in annualized revenue and several hundred percent growth, he says, venture investors continued to pass. Daimler approached with a possible acquisition around €100 million; Villig, who recalls owning about 45%, wanted to keep building. His co-founders supported him and the offer was declined.

About a year later, after the business grew substantially, Daimler invested more than $100 million at a $1 billion valuation, according to Villig. Additional investors brought that round to around $170 million. He says the team retained its cost-conscious culture while using the capital to enter more cities and fund the expensive early marketplace phase.

Launching a major city can require very different sums from a small Baltic market. Villig estimates that reaching critical mass in London takes at least €100 million; he contrasts that with Bolt’s initial $5,000 launch in Estonia. These are his rough estimates, not comparable audited launch costs.

He describes Bolt’s growth as approximately zero to $2 million ARR in 18 months, then $10 million in the next 18 months, then $100 million in under two years, followed by around $2 billion ARR several years later. ARR is an annualized run rate, not necessarily recognized revenue, and the interview does not independently verify the figures.

Villig says New York investors such as D1 and TCV engaged deeply with numbers, market shares and trends, while earlier venture investors focused more on the category narrative. He reports about $1.5 billion raised in total, with roughly 10% from strategic investors. These descriptions of investor styles are his impressions.

When COVID-19 hit, he says Bolt’s approximately $200 million annualized revenue fell 85% in four weeks. Management chose not to lay off staff and instead reduced salaries by 20%, with some employees voluntarily accepting larger cuts. The decision assumed a recovery within six months and was a significant gamble, he says.

The company reviewed costs and recruited drivers online while travel demand was depressed. As individual cities reopened, a team monitored them closely and spent on marketing and discounts. Villig says revenue recovered in about five or six months and market share in some places doubled or tripled; these are his company’s claims.

Sequoia approached during a later round and, Villig recalls, made a decision within 48 hours after two calls. He says the 2021 round valued the company at roughly $4.5 billion. In his view, Sequoia’s brand helped with recruiting, even though Bolt already had an operating model he believed could scale.

Villig says he was poor at hiring initially and fired seven of his first ten hires. He attributes this to optimistic assumptions and an inadequate vetting process, then says Bolt codified its hiring. He values intelligence and decision-making, citing the technical co-founder and the head of growth as examples of people who had outsized impact.

He agrees that Europe has relatively few leaders who have scaled technology organizations to thousands of employees, but thinks strong engineering and marketing talent is available. Bolt’s experience, he says, has favored talented people from Central and Eastern Europe who could grow with the company.

Villig says European venture funding has improved but still lacks the depth of US pension and endowment capital. Stebbings disagrees, arguing too much money chases too few obvious opportunities and inflates prices. Villig replies that insufficient funding can itself prevent capable companies from competing, and asks investors to tolerate longer paths to profitability.

He advises a European founder to consider European investors for local recruiting help and US investors when the target market is the US. He estimates most investors primarily provide capital and says he would have approached New York funds earlier. These are his subjective views as a founder seeking capital.

Bolt’s expansion beyond ride-hailing followed its ambition to offer alternatives to private car ownership. Around 2018, with enough resources for another bet, it chose electric scooters. Employees and investors worried about hardware difficulty and cannibalizing short car trips; Villig believed offering customers another option was better than leaving the opportunity to a competitor.

The first scooter launch in Paris was difficult. Villig says vandalism and theft at one point removed about 3% of vehicles each week, making unit economics unworkable. Bolt tried the scooters in Estonia instead, where usage and economics were stronger than expected.

He estimates good micromobility markets can reach 20% contribution margins while poor markets can be negative. That is a contribution-margin measure, not company-wide net profit. The business also needs warehouses, charging, maintenance and recovery, and he says more than a thousand people work on the operation.

Off-the-shelf scooters broke quickly because they were not designed for sharing, he says. Bolt recruited an Estonian team of around ten electric-vehicle engineers who had manufacturing connections in China. They designed successive generations of hardware, and Villig estimates the effort ultimately used around 10% of total funding.

On self-driving, Villig expects major long-term change but says commercially viable systems cheaper than human drivers and approved by regulators are still years away. He distinguishes end-to-end approaches from sensor-heavy systems and argues that a rider’s fare does not reveal the provider’s underlying cost. His comparisons and timelines are forecasts from a mobility-platform CEO, not independent safety or cost findings.

He hopes technology and costs improve within five years, with regulation potentially taking longer, particularly in Europe. Villig expects autonomous technology providers to partner with operators such as Bolt for fleet operations, local compliance, payments, support, cleaning and charging. That prediction supports Bolt’s own strategic interests.

Asked about Uber, he says competitors have raised much more capital. He says some Bolt markets are profitable and generate cash to invest elsewhere, but growth remains the priority and he does not claim the whole company is already consistently profitable or paying dividends.

Bolt operates ride-hailing, scooters and e-bikes, car rental, restaurant delivery and grocery delivery, he says. Car rental is currently the smallest by revenue, but he expects it to become a multibillion-dollar category and is particularly optimistic about on-demand cars in cities. These are ambitions, not achieved revenue.

Among possible future products he is interested in dining payments: customers could scan a code, order and pay, linked to Bolt’s food and loyalty ecosystem. Stebbings says he has found scan-to-pay a difficult business. Villig believes Bolt can improve the economics through cost efficiency.

Villig’s biggest strategic regret is not continuing faster expansion after larger funding rounds. He says advice to adopt more process led Bolt to slow launches and focus on existing markets, leaving opportunities for competitors. That is his retrospective assessment, not proof a faster approach would have succeeded.

He repeats his contrarian view that ride-hailing platforms are well placed to distribute autonomous cars because operating large fleets is difficult. He says money is important up to the point of meeting basic needs, but his greatest satisfaction comes from seeing the company and its people grow.

Initially skeptical about generative AI, Villig says his view changed after using language models regularly and seeing potential beyond customer support, including internal information retrieval, document summaries and sales tools. These are anticipated uses rather than measured improvements.

He names war, particularly Russia’s attack on Ukraine, as a major concern. For health he recommends basic habits such as sleep, exercise and time with friends, and says he has avoided sacrificing sleep during Bolt’s development. This is personal advice and experience, not a clinical prescription.

Finally, Villig argues that many on-demand markets tend toward two major players and that acquisitions aimed at monopoly returns often fail to prevent a new second competitor emerging. He also thinks high entry barriers now make traditional new ride-hailing entrants less likely, with future disruption more likely to come from new transport modes or autonomous vehicles. These are his market theories and forecasts.