Inside Rocket Internet’s troubled start-up factory

An in-depth examination of Rocket Internet’s operations reveals an incubator that’s much better at starting new companies than running them


Oliver Samwer, chief executive officer of Rocket Internet SE, during an interview at the company’s new headquarters in Berlin, Germany. Photo: Bloomberg
Oliver Samwer, chief executive officer of Rocket Internet SE, during an interview at the company’s new headquarters in Berlin, Germany. Photo: Bloomberg

London/Berlin/Frankfurt: In November 2015, Oliver Samwer had a new investment to pitch. The charismatic founder of Rocket Internet, a Berlin-based incubator famous for churning out copycat start-ups based on successful US tech companies, was touting his latest baby, a kind of Airbnb for college undergrads called Nestpick. Samwer’s team invited Luxembourg venture firm Mangrove Capital Partners to co-invest and extolled Nestpick’s growth potential and the power of Rocket’s globe-girdling start-up machine.

The pitch slid down easy as a third martini, and Mangrove agreed to take part in an $11 million funding round. The hangover wasn’t long in coming. Less than six months later, Nestpick was struggling and had fired most of its 140 employees. Rocket, it seems, had misjudged the basics of the student housing market, where demand spikes just before school starts and evaporates during the academic year. Today, Mangrove chief executive officer Mark Tluszcz regrets getting involved with Rocket and says Nestpick’s survival is far from assured. “You’re super-excited the day you put money in,“ he says. “But within a year you wonder what you were thinking.“

Investors feel much the same about Rocket these days. When the German incubator went public on the Frankfurt stock exchange in October 2014, it was Germany’s biggest IPO in about seven years and the first chance public investors had to own a stake in some of the fastest growing start-ups in Europe. The listing added more than $2 billion to Rocket’s coffers, valuing the company at $8.2 billion. Rocket’s stock price has since cratered, falling 53%. Analysts criticize the company’s opaque financial reporting and question its ability to turn a profit or find lucrative exits for its start-ups.

Rocket can point to some successes. Among them: a Zappos-like firm called Zalando that is now one of Europe’s largest fashion retailers; CityDeal, a coupon clone sold to Groupon in 2010; and Lazada, a South-east Asia-focused consumer electronics e-commerce company sold to Alibaba earlier this year. The company has also helped spawn start-up hubs—from hometown Berlin to emerging markets around the world. “Rocket is in part responsible for the fact that Berlin is now a tech hotbed,“ says Sarah Simon, an analyst at Berenberg Bank in London. “They’ve incubated companies and people, and they succeeded in raising an enormous amount of capital—which people tend to forget.“

All the same, most Rocket start-ups have struggled, with at least a dozen closing and almost all losing money. On 2 September, the company issued its first profit warning, telling investors it had lost €617 million ($707 million) in the first half. The caution came after Rocket slashed the value of e-retailer Global Fashion Group, a company Rocket had called a “proven winner,” by two-thirds to about €1 billion. A week later, Rocket announced a funding round for another “proven winner,” online furniture retailer Home24, which cut its value by more than 50%.

Also Read: The rise and fall of Rocket Internet in India

When Rocket announced first-half results on 22 September, it became clear that the company was paring its losses at many start-ups, news that lifted the shares from an all-time low. But though a majority of analysts maintain a buy rating on the stock, there remain reasons for concern: HelloFresh, a grocery-box delivery business and Rocket’s single-largest investment, saw growth stutter in the second quarter and losses more than double. Rocket’s businesses in India and Africa, once heralded as big growth markets for the incubator, are being sold for pennies on the dollar, firing staff or closing completely. The company’s attempt to expand into financial technology has also floundered, with an online payments copycat it backed called Paymill filing for insolvency earlier this year.

Samwer has pledged to stanch the red ink and says at least three of his start-ups–which ones he won’t say—will be profitable by the end of 2017. He has also applied to list Rocket shares on the portion of Germany’s stock exchange reserved for established, big-cap companies, which requires extra transparency and more frequent financial reporting. But with many investors feeling burned, Rocket’s launch window may be closing. An examination of Rocket’s history and current operations—including interviews with more than a dozen former and current employees—reveal a start-up factory that is much better at launching companies than actually running them.

"The performances remind him of Steve Jobs’s famous reality distortion field"- Former Rocket executive

Oliver Samwer is sitting in a conference room on the 16th floor of Rocket’s new headquarters, a tower located in the German capital’s trendy Kreuzberg district. Now 44, his wavy brown hair graying, Samwer appears ever-so-slightly disheveled: the tails of his blue dress shirt hang untucked. But if the recent criticism or ailing stock price trouble him, it’s not apparent. Samwer still has the gracious manner and winning smile of the wunderkind who wowed Germany in 1999 when he became the country’s first Internet millionaire with an EBay knock-off he and his brothers started and sold to EBay just 100 days later for $53 million.

During a 45-minute interview, Samwer defends Rocket against those who criticize the company for having few successful IPOs since its own listing in 2014. “Exits are not our focus,” he says. “Our focus is to build companies and keep them for a long period of time.” Rocket should not be confused with a venture capital firm–which is primarily a financial investor, he says. “We’re an operational company that basically starts from scratch and invests in companies and makes them through operational infrastructure—the people, the processes, the data–better.”

Brokers work during the IPO of German start-up company Rocket Internet at the stock exchange in Frankfurt, Germany. Photo: AFP
Brokers work during the IPO of German start-up company Rocket Internet at the stock exchange in Frankfurt, Germany. Photo: AFP

It’s a classic performance from a charming and shrewd salesman. Samwer constantly rehearses his pitches and sweats the details right down to the food served at investor meetings, according to venture capitalists and former Rocket employees. One former Rocket executive says the performances remind him of Steve Jobs’s famous reality distortion field—when inconvenient facts rarely get in the way of a good story. Several people who have worked with Samwer say he pushes start-ups to emphasize data that puts them in the best possible light—like using monthly user figures when the start-up sells subscriptions by the week.

Two former Rocket executives express amazement that sophisticated investors fail to ask obvious questions following Samwer presentations. Some decided to invest on the spot, only to sheepishly return weeks later to ask mandatory due diligence questions. Samwer tells Bloomberg that he tries to convey data demonstrating an industry’s growth potential and Rocket’s expertise building companies. “I think people feel the experience in what we’re doing, the belief in what we’re doing and the rest is up to themselves,” he says. “There’s no special magic.”

Hailing from a well-to-do Cologne family, Samwer and his two brothers, Alex and Marc, honed their approach to entrepreneurship over the course of two decades of investing. After their initial success with Zalando, the EBay clone, they founded a ringtone business called Jamba!, which they sold to VeriSign, a network infrastructure company, for $273 million in 2004. They also achieved success with German knock-offs of YouTube, Twitter and Facebook. (They made some savvy investments in the originals too, profiting from bets on Facebook, LinkedIn and Zynga.)

In 2007, they founded Rocket as a clone factory, racing to launch knock-off start-ups in markets from Western Europe to Latin America to Africa before the originals arrived. As befitted Rocket’s name, speed was critical–growing companies at breakneck rates in order to lock in first-mover advantage. Silicon Valley CEOs and venture capitalists considered the German incubator a parasite, leaching off American creativity. Germans, justly proud of their nation’s engineering and entrepreneurial prowess, were ambivalent about having Samwer as their Internet posterboy. He defends the copycat approach, noting that many of the world’s most successful businesses—from General Electric to Toyota—never invented a new business model; they just executed better. “Compared to most others,“ he says, “we can provide operational value.”

A Rocket team is constantly on the lookout for business ideas, scouring the US tech press and funding round announcements from prominent Valley venture firms. The target identified, a smaller group studies potential markets and devises a business plan. When a laundry delivery service called Washio was gaining traction in New York, Rocket assigned a team of young employees to consider cloning the business in London. They called 20 area laundry mats to gauge pricing, quizzed logistics firms about handling deliveries and investigated the cost of leasing a warehouse and filling it with washing machines.

Once Rocket greenlights an idea, the company throws staff at it–assigning them from a central pool of managers as well as its own marketing, engineering and accounting departments. It hires legions of young, ambitious people, many from university business programs, top-tier consulting firms or investment banks. For the laundry business, which Rocket dubbed ZipJet, the company installed former McKinsey and Boston Consulting Group consultants as co-founders and launched it in London, Berlin and Paris. ZipJet still exists, though Washio, the original New York company, has since folded.

"Rocket has a surprisingly hard time keeping people"

This thundering herd approach allows Rocket to ramp up in weeks where others would take months. At first, all the workers assigned to a project come from Rocket, but as a start-up grows it’s expected to hire its own staff; returning Rocket employees are then assigned to new projects. Start-ups are also eventually expected to find their own office space. In the meantime, Rocket charges its portfolio companies management fees to cover the salaries of the staff and real estate costs, a practice common to private equity but unusual in venture capital.

Rocket start-ups can track more than 800 metrics, from Facebook likes to order return rates. Samwer says he combines a variety of metrics into three yardsticks he calls “contribution margin one, two and three.” These relate to gross profit, logistics costs and customer acquisition. By looking at ratios between these measures—for instance comparing customer acquisition rates with the expected lifetime value of each customer—Samwer decides when he wants to hit the gas on an investment, or slam on the brakes. Samwer says traditional financial measures—like revenue and ebitda—don’t always work for start-ups. If you looked only at these, he says, “you’d probably never start a business because it only looks like losses.”

Some of those who have worked at Rocket, though, say the machine-like system for ramping up start-ups favours speed over stability. Home24, the furniture-delivery business, has been plagued by logistical problems that may have played a role in its recent valuation write-down. In one example, returns spiked because buggy programming resulted in the company’s website sending incorrect order information to the company’s shipping department. It often took weeks to discover the problem, according to a person familiar with the mistakes. Home24 tells Bloomberg it has since fixed these issues. Meanwhile, HelloFresh, the food-delivery business, put off a planned IPO as losses piled up and instead cut back on spending for customer service and warehouse staffing to help bend itself towards profitability, another person said.

Samwer says one of Rocket’s key strengths is its ability to leverage institutional knowledge across its portfolio—taking lessons from one start-up and applying them to the rest. “If someone wants to do last-mile logistics in a certain country,“ he says, “we can find him the person who has done it before.” But a half-dozen former Rocket employees say young staffers were put in charge of functions with which they had no experience. At HelloFresh, two of these people say, managers with no background in food or logistics were asked to negotiate with food wholesalers, rent warehouse space and organize delivery procedures. Most of the former Rocket employees interviewed also said there were few formal mechanisms to transfer know-how between Rocket portfolio companies so start-ups made the same mistakes over and over.

Consider Bonativo, an organic grocery delivery service Rocket launched in January 2015 and rapidly expanded to Berlin, Hamburg, London and Amsterdam. Its young managers, according to a former senior executive, failed to absorb lessons from Zalando about checkout procedures and how to set up distribution networks. Unable to make headway against entrenched competitors, Bonativo lasted 17 months and closed in June. In the prospectus for its most recent share offering, Rocket noted that “the executive officers of the companies in the Rocket Internet network often have limited management experience in their respective industries.”

"This thundering herd approach allows Rocket to ramp up in weeks"

For a company that sells investors on a pool of managerial talent that can be deployed at will, Rocket has a surprisingly hard time keeping people. Many recruits use the company as a CV booster for a year or two and then decamp for a non-Rocket start-up. Long hours and a demanding environment partly explain high turnover. Several former Rocket senior managers say Samwer can be a micromanaging and domineering boss, calling start-up founders at all hours to pepper them with questions. He can be harsh, too. Two former Rocket executives describe a meeting in London when Samwer bullied two start-up founders over a television spot he deemed unworthy. “This is shit,” he said. “Stop trying to be Hollywood.” Then he asked the founders where they went to school. “And I thought you were supposed to be smart,” he said. Samwer, who is also accused of shouting and swearing at subordinates, says he has mellowed with age. “The older I get, I’m less impatient,” he says. “Should one raise one’s voice? No. No doubt about it. But we are a passionate firm.”

A mellower Samwer probably won’t halt the brain drain, however. The biggest complaint from employees present and former is that Rocket doesn’t compensate its people adequately. The prospect of becoming an instant millionaire buys a lot of loyalty even under the harshest and most frantic conditions. But unlike most start-ups, Rocket companies don’t hand out shares to employees. Most receive no equity at all; even “founders” typically get a 3% equity stake. At HelloFresh, the ready-to-cook meal delivery service, a dozen or so staffers out of roughly 1,000 employees globally received small equity stakes . In lieu of stock, employees often receive fancy titles (entrepreneur-in-residence, co-founder) and salaries only slightly better than they could pull down at any European start-up. Venture firms prefer start-up founders to have skin the game. “You have these founders who are not really founders, but hired guns,” says Tluszcz, the Mangrove partner. “That’s not our style. We want the founders to have more equity than anyone else.”

It doesn’t help that Samwer has alienated experienced managers by breaking promises about their compensation, particularly equity stakes, according to two former Rocket executives. He says it got to the point where he didn’t trust anything Samwer said unless it was written down and notarized. Samwer denies this. “It is not correct,” he says, adding that some founders, if they were fired or resigned for under performance, might lose equity rights. “I do not believe that came from people who have been long and successful with us,” he says. He also says Rocket has created more than 100 millionaires–“more than anybody else in the European Internet economy”–mostly through the Zalando IPO; Samwer didn’t specify any other Rocket company that he believes generated this kind of widespread wealth. Few Rocket employees seem to have walked away with the sort of windfall the Europe-based team that built Skype received, for instance. Nor does Berlin buzz with tales of receptionists and mail room clerks for Rocket becoming millionaires as happened with Facebook, PayPal or Google in Silicon Valley.

The four original Rocket managing directors who were influential in the company’s initial success–Florian Heinemann, Uwe Horstmann, Felix Jahn and Christian Weiss–have all departed, often after reported disagreements with the Samwer brothers over compensation. All four are now active as venture capitalists, often backing companies that compete with Rocket start-ups. Rocket, in an e-mail to Bloomberg, declined to comment on the executives’ departures other than to say each was “an individual decision.“ More recently, in January 2016, Rocket lost its general counsel and a senior finance executive. A large portion of Rocket’s public relations team has also turned over in the past year. Samwer says Rocket’s staff turnover is not unusually high and he shrugs off the idea that defections have hurt the incubator. Having managers leave “is part of building a big company,” he says. And he noted that the founding teams at HelloFresh, GFG, Lazada and Jumia “are still the same ones that started it.”

Also Read: The rise of the unicorns

For several years, there was no more ardent Rocket supporter than Kinnevik, a Swedish asset management firm based in Stockholm. The relationship began after Samwer cultivated a relationship with Kinneviks’s principal owner Cristina Stenbeck, heiress to an industrial fortune. The fund invested €631 million in Zalando and €155 million in Rocket, becoming the biggest shareholder after Samwer and his brothers, and still owns a 13% stake. The Swedish fund also co-invested in many Rocket start-ups.

But today that once close relationship has frayed. The precipitating event, according to people familiar with the situation, was a $742 million venture fund Samwer started last year to co-invest in late-stage technology companies alongside Rocket. Kinnevik CEO Lorenzo Grabau told Samwer the fund would present a conflict since Kinnevik also invests in many late-stage tech firms. When Samwer went ahead anyway, Grabau resigned as Rocket chairman, though he remained on the board through the spring. He tells Bloomberg that Kinnevik’s five-year co-investment deal with Rocket had been successful but that both sides have chosen not to renew it. “We’ve ended the partnership,” he says. “We’ve built some great companies. And we’ve each gone our own way.” Kinnevik, however, continues to do follow-on investments in Rocket start-ups in which it already owns a stake. “We have a very good relationship with Kinnevik,“ Rocket said in an e-mailed response to questions about its partnership with the Swedish investor.

Even before the falling-out, Kinnevik’s investments in Rocket companies brought attention to an issue that has been giving analysts pause for a while: how Rocket values its start-ups. The company uses something called last portfolio value, or LPV, which calculates value based on a start-up’s most recent financing round. The methodology is controversial, says Northern Trust Securities analyst Neil Campling, because Rocket usually participates in these funding rounds and so is in a position to inflate a given valuation. He also says LPVs may reflect market conditions that are months or years out of date.

Using more traditional methods, Kinnevik calculated much lower valuations for Rocket companies. It listed Westwing, a home goods e-commerce company, at an implied value of €254 million as of 30 June. Rocket lists the company on its books at €480 million, based on its last funding round in March 2015. (Kinnevik’s valuations include the effect of contract terms—known as liquidation preferences—that later investors in start-ups often insist upon. As a result, implied valuations based on Kinnevik’s assessment of its stakes in these companies may underestimate the start-up’s overall enterprise value. But, notably, Rocket’s own LPV methodology doesn’t take liquidation preferences into account even though it too may be subject to these same contract terms.) Rocket values furniture merchant Home24 at €420—seven times the valuation implied by Kinnevik metrics. Kinnevik and Rocket reportedly clashed over HelloFresh’s planned IPO, with Samwer pushing for a €3.3 billion valuation—far more than Kinnevik thought realistic. (Both sides deny a disagreement prompted to them pull the listing.)

Also Read: The trouble with Foodpanda

In recent weeks, Rocket has written down the valuations of several of its companies, further spooking investors. The company is now talking up the prospects of restaurant delivery service Delivery Hero, another “proven winner“ and a potential IPO candidate in the coming months. Yet analysts are still struggling to get a fix on the start-up’s finances. Northern Trust’s Campling points to a chart in Rocket’s March earnings release that showed improving profit margins for the company’s proven winners. It excluded Delivery Hero. So did Rocket’s September results. Samwer says accounting rules won’t permit Rocket to include figures for Delivery Hero; the start-up’s chief says the company will break even by year-end. But a top former Rocket manager fears Samwer is pursuing a “lucky punch“ strategy—using a Delivery Hero listing to silence critics and pacify investors. This person says an IPO could buy Rocket time if it does well—or hurt the incubator if it performs poorly.

Samwer brushes aside such concerns, saying Rocket is in better shape than it’s ever been. He points to €1.7 billion in cash, a new co-investment fund that lets the company back its start-ups through later stage of growth, and Rocket’s vast footprint, with 30,000 employees in 120 countries. Grabau says investors have continued to support Rocket and Samwer’s new investment vehicle because there are so few options for tech investment in Europe. “If you want to build a portfolio of 10 Internet companies in Europe,“ he says, “you very quickly run out of names.”

But the ground is shifting. Rocket once dominated Berlin’s tech hub. Now it competes with start-ups backed by VC firms from elsewhere in Europe and Silicon Valley. There are thriving tech hubs in other European cities too, not to mention the emerging markets where Rocket has invested heavily. And many of the start-ups in these places are pioneering cutting-edge technology and new business models, not just aping those from the US. “The day and age of copying seems over,” says Ciaran O’Leary, a general partner at Berlin-based venture firm BlueYard Capital.

To the long list of critics who dislike Samwer’s approach, Rocket’s recent stumbles seem poetic justice. But in his perch atop Rocket’s new office tower, even Samwer seems to have changed his emphasis. These days, Rocket is less about starting new copycat businesses, he says, and more about ensuring the existing ones make money. “If you look at the underlying companies, we’re making tremendous progress,” he says. Rocket never guaranteed investors a quick buck, he adds, noting that the prospectus for its IPO talked of a six- to nine-year horizon. “We never made the promise, ‘we are, one year after IPO, profitable.’” Samwer says he’s learned to be more patient. Now he must persuade investors do the same. Bloomberg

READ MORE