Connect with us

Tech News

Measuring the full impact of digital capital

Published

on

Although largely uncounted, intangible digital assets may hold an important key to understanding competition and growth in the Internet era.

July 2013 | by Jacques Bughin and James Manyika

digital-assets

On July 31, 2013, the US Bureau of Economic Analysis released, for the first time, GDP figures categorizing research and development as fixed investment. It will join software in a new category called intellectual-property products.

In our knowledge-based economy, this is a sensible move that brings GDP accounting closer to economic reality. And while that may seem like an arcane shift relevant only to a small number of economists, the need for the change reflects a broader mismatch between our digital economy and the way we account for it. This problem has serious top-management implications.

To understand the mismatch, you need to understand what we call digital capital—the resources behind the processes key to developing new products and services for the digital economy. Digital capital takes two forms. The first is traditionally counted tangible assets, such as servers, routers, online-purchasing platforms, and basic Internet software. They appear as capital investment on company books. Yet a large and growing portion of what’s powering today’s digital economy consists of a second type of digital capital—intangible assets.

They are manifold: the unique designs that engage large numbers of users and improve their digital experiences; the digital capture of user behavior, contributions, and social profiles; the environments that encourage consumers to access products and services; and the intense big-data and analytics capabilities that can guide operations and business growth. They also include a growing range of new business models for monetizing digital activity, such as patents and processes that can be licensed for royalty income, and the brand equity that companies like Google or Amazon.com create through digital engagement.

Conventional accounting treats these capabilities not as company investments but as expenses, which means that their funding isn’t reflected as capital. Since the amounts spent aren’t amortized, they take a large bite out of reported income. Spending on those capabilities sometimes should be treated as capital, though, since they can be long-lived. Amazon.com’s development of an internal search process that promotes recurring sales or the efforts of Netflix to fine-tune personal recommendations to increase video viewing and retain customers are certainly more than expenses. Such capabilities, which are complex to build and replicate, can often help companies create enduring competitive strengths.

We’re acutely aware of misguided efforts to justify sky-high valuations during the late-1990s Internet bubble by claiming that finance and accounting fundamentals were no longer relevant. We also recognize that we’re far from the first to note the relationship among intangibles, company-level growth and productivity, and overall economic growth.1 What we want to suggest here is that those relationships, which once represented a small minority of business activities, are becoming the rule in the digital economy. In fact, much of today’s digital spending could pay for long-lived intangible assets that will define the competitive landscape going forward.2 The rising stakes are seen in the copyright battles between Internet and consumer-electronics companies and in major spending on patent portfolios.

Above all, we want to emphasize the importance, for many business leaders, of making the mind-set shift required to embrace the importance of digital capital fully. The disruptive nature of digital assets is intensifying in markets such as search, e-commerce, and social media (where attackers can build business models with near-limitless scale). Disruptive digital assets are also important in segments where behavioral data and user participation can be monetized, by defining entirely new business opportunities or fostering breakthroughs in collaborative innovation. As the mobile-payments start-up Square is demonstrating in the credit-card arena, increasingly, companies that deploy these assets have the potential to threaten large existing profit pools thanks to the challengers’ vastly different economics or radically new ways of doing things.

The big picture

There are parallels between what’s occurring today and during the period, 100 years ago, when electric motors gained widespread adoption. Early in that cycle, companies invested in physical motors, which like today’s servers and routers provided a new growth platform. But the more important kind of value appeared after companies began to understand how motors could change almost every process, improve productivity, and stimulate innovation. Companies that captured these benefits were more successful and more valuable than others.

Today, the market valuations of many Internet-based companies are higher than those of their counterparts in other sectors, including high tech. Many Internet leaders earn lower returns on equity than established technology companies do, yet there’s no reason to believe that markets are making irrational bets on the growth potential of digitally adept companies. As the sidebar “Valuation and intangibles: Viewing the numbers differently” illustrates, treating digital intangibles as assets rather than expenses clarifies the logic behind valuations. (We based these pro-forma valuation calculations on data compiled by academic researchers, as well as assumptions about rates of intangible and digital investment from our own and outside research.)

Macroeconomic studies we have done suggest that digital capital is not only growing rapidly but has also become a major contributing factor in global economic growth.3 We examined the national-accounts data of 40 countries, assigning values to tangible and intangible assets. In 2005, digital-capital investment represented barely 0.8 percent of GDP for those countries. This year, it will exceed 3.1 percent of GDP. Likewise, the accumulating global value of digital-capital investments has reached more than $6 trillion, about 8.5 percent of nominal world GDP. Globally, levels of digital intangible investment are more than half those of digital tangible investment. In more highly digitized economies, such as Israel, Japan, Sweden, the United Kingdom, and the United States, spending on intangibles represents two-thirds of digital capital’s total value.

This activity is starting to power growth. We estimate that digital capital is the source of more than one percentage point of global GDP growth (roughly one-third of total growth). Intangible capital already accounts for two-thirds of that slice, tangible investment for the rest. This growth flows from not only capital deepening but also increased labor productivity—a remarkable thing, since the digital economy has emerged in the relatively brief space of 15 years. By contrast, it took 80 years for steam engines to increase labor productivity to the same extent, about 40 for electricity, and more than 20 for conventional information and communications technologies.4 (For more on the relationship between capital formation and productivity, see sidebar “Innovation, capital, and productivity growth.”)

Navigating the new terrain

Intangible digital capital’s role in economic growth gives policy makers one more reason to favor investments in broadband and other forms of Internet infrastructure. Such investments correlate strongly with overall digital-capital levels. In our experience, though, the implications are even greater for executives, who often are not tuned into their organizations’ digital strengths or weakness. Few companies have gone through the internal exercise of reclassifying expenditures or segregating benefits from spending on intangibles. And of course, companies can boast a high ROE thanks to strong legacy-product margins but may nonetheless have muted growth prospects as a result of underinvesting in digital capital. To set a more effective digital course, leaders should consider the following ideas.

Take stock of your assets

Since identifying intangible assets is difficult, companies may be missing growth opportunities. Many have realized only recently that they can use social-media interactions with their best customers to leverage innovation efforts or that they may have unused data they could restructure into valuable big-data assets to sharpen business strategy. Similarly, companies should take stock of how digital capital they don’t own may be relevant to the business. A retailer that doesn’t have access to digital behavioral data on consumers, for example, may be at a disadvantage. So could a bank whose customers access products through a third-party platform that limits the bank’s ability to capture information.

Conversely, companies may wrongly assume that their growth results from conventional capital spending and therefore compromise growth by underinvesting in digital competencies. One online company, for example, stuck to a subscriber pay model in hopes of boosting returns on tangible investments such as server farms. It wound up missing a massive social-networking opportunity that would have yielded far greater returns on advertising revenues.

Our global research shows that the stock of intangible assets varies considerably by region. Some markets have larger numbers of strong digital contenders, others fewer. Companies could make those differences a factor in deciding which markets to enter and where to place digital bets.

Face up to looming threats

Assume that digital leaders in your competitive zone are relentlessly expanding their intangible assets both to attack existing markets and to create new ones. Amazon.com, for instance, won share from brick-and-mortar retailers with its ease-of-purchase model and its ability to reach long-tail customers. Now it’s launching new business models (such as Amazon Prime) to further leverage its user base and logistics capabilities. It’s also using tangible server assets to offer cloud-based labor services (Mechanical Turk) that match freelance workers with demand for their labor.

A good first step is to identify which areas of your value chain are most vulnerable—for example, service delivery or weak digital brands. Competitors can slide vertically or horizontally into large gaps, so you’ll need to build digital assets quickly as a counterweight. Even companies that have a considerable stock of digital assets should understand that capturing value from them isn’t a given. Instead, such companies must define (and relentlessly innovate with) business models that can be scaled up to match those assets.

One clue suggesting that a company might face emerging digital challenges is the existence of businesses that have unusually high levels of revenue per employee in adjacent market spaces. Amazon.com’s employee productivity, for example, is double that of traditional retailers. Netflix, similarly, generates more revenue per employee than traditional cable operators do, by leveraging intangibles such as its highly evolved recommendation algorithms. Unusual financial profiles are another warning sign. Since digital funding is counted as operating expenditure, digital leaders often have small capital-investment levels relative to their size and growth potential. They also borrow less, both because they may not need to (some reap sizable market rents from, for example, search licensing fees or patent income) and because banks may be less likely to lend against intangible assets.

Partner with care

Most companies rely on digital agencies for things like optimizing search marketing. In such cases, they may be ceding digital capital, since they never develop a full understanding of consumer segments or what inspires a customer who searches for their products. Seeing such capability building as an investment may change the logic of using third parties. Similarly, when companies look to established tech players for partnerships shoring up weaknesses, they should be cautious: some seemingly high performers may be on the wrong path and could burden you with outmoded standards and platforms. Alternatively, if you deal with strong players, you may be leaving yourself vulnerable by letting them lead.

The need for growth and competitiveness will force companies to build strong digital capabilities. Viewing them as assets rather than additional areas of spending requires a new set of management and financial lenses. Embracing them is a major shift—but one worth making for companies striving to master a still-evolving landscape.

About the authors

Jacques Bughin is a director in McKinsey’s Brussels office. James Manyika is a director of the McKinsey Global Institute and a director in the San Francisco office.

Continue Reading
Click to comment

Leave a Reply

This site uses Akismet to reduce spam. Learn how your comment data is processed.

Business

How McKinsey Lost Its Way in South Africa – a special report

Published

on

By

The McKinsey & Company offices in Sandton, Johannesburg’s financial center. In 2015, the consulting giant entered into a contract that turned out to be illegal.CreditGulshan Khan for The New York Times

JOHANNESBURG — The blackouts kept coming. The state-owned power company, Eskom, was on the verge of insolvency. Maintenance was being deferred. And a major boiler exploded, threatening the national grid.

McKinsey & Company, the godfather of management consulting, thought it could help, but was not sure that it should, according to people involved in the debate. The risk was huge. Could McKinsey fix the problems? Would it get paid? Would it be tainted by South Africa’s rampant political corruption?

In late 2015, over objections from at least three influential McKinsey partners, the firm decided the risk was worth taking and signed on to what would become its biggest contract ever in Africa, with a potential value of $700 million.

It was also the biggest mistake in McKinsey’s nine-decade history.

The contract turned out to be illegal, a violation of South African contracting law, with some of the payments channeled to an associate of an Indian-born family, the Guptas, at the center of a swirling corruption scandal. Then there was the lavish size of that payout. It did not take a Harvard Business School graduate to explain why South Africans might get angry seeing a wealthy American firm cart away so much public money in a country with the worst income inequality in the world and a youth unemployment rate over 50 percent.

And a bitter irony: While McKinsey’s pay was supposed to be based entirely on its results, it is far from clear that the flailing power company is much better off than it was before.

The Eskom affair is now part of an expansive investigation by South African authorities into how the Guptas used their friendships with Jacob Zuma, then the country’s president, and his son to manipulate and control state-owned enterprises for personal gain. International corruption watchdogs call it a case of “state capture.” Lawmakers here call it a silent coup. It has already led to Mr. Zuma’s ouster and a moment of reckoning for post-apartheid South Africa.

Yet despite extensive coverage of the scandal by the local news media, one question has remained largely unanswered: How did McKinsey, with its vast influence, impeccable research credentials and record of advising companies and governments on best practices, become entangled in such an untoward affair?

McKinsey admits errors in judgment while denying any illegality. Two senior partners, the firm says, bear most of the blame for what went wrong. But an investigation by The New York Times, including interviews with 16 current and former partners, found that the roots of the problem go deeper — to a changing corporate culture that opened the way for an aggressive push into more government consulting, as well as new methods of compensation. While the changes helped McKinsey nearly double in size over the last decade, they introduced more reputational risk.

The firm also missed warning signs about the possible involvement of the Guptas, and only belatedly realized the insufficiency of its risk management for state-owned companies. Supervisors who might have vetoed or modified the contract were not South African and lacked the local knowledge to sense trouble ahead. And having poorly vetted its subcontractor, McKinsey was less than forthcoming when asked to explain its role in the emerging scandal.

“I take responsibility,” said Dominic Barton, McKinsey’s managing director, who is stepping down at the end of June as previously planned.CreditSasha Maslov for The New York Times

Since the Eskom disclosures, much of McKinsey’s business in South Africa has evaporated. Mr. Barton has made six trips there to assess the damage and make amends, and McKinsey has asked its 2,000 global partners to repay South Africa, where it is under investigation.

Indeed, the harm to the McKinsey brand is more profound than the fallout from the epochal Galleon hedge fund case almost a decade ago, in which McKinsey’s former managing director and a senior partner were convicted on charges related to insider trading. Neither man acted on behalf of McKinsey.

More broadly, the scandal in South Africa — which has ensnared several other overseas companies — underscores the risks that arise as governments increasingly turn over responsibilities to consultants who operate mostly in secret, with little or no public accountability.

McKinsey built its brand as the ubiquitous adviser to businesses great and small. But in recent years, it has created an increasingly powerful unseen presence as counselor to governments across the globe.

The extent of that global influence is difficult to evaluate because, as a matter of policy, the firm will not reveal clients or the advice it gives.

Even so, by examining government records, along with McKinsey publications and other company documents, The Times found that the firm shapes everything from education, transportation, energy and medical care to the restructuring of economies and the fighting of wars.

McKinsey’s clients include sovereign wealth funds worth more than a trillion dollars, as well as what one marketing brochure describes as “defense ministries, military forces, police forces and justice ministries in 15 countries,” where the company consults on such matters as the maintenance and support of “armored personnel carriers; minesweepers, destroyers and submarines; and fast jets and transport aircraft.”

McKinsey “is a hidden, unaccountable power that has a prestigious face,” said Janine R. Wedel, a professor at George Mason University who has written extensively on what she calls “the shadow elite.” She added, “Think of them as a repository of the most intimate information that governments and others have, from what they are investing in to who wields influence.”

McKinsey refused to work in South Africa until it embraced democracy in the mid-1990s, but records show that it consults for many authoritarian governments, including the world’s mightiest, China, to a degree unheard of for a foreign company. Late last year, two McKinsey partners spoke at a meeting of the state-controlled conglomerate China Merchants Group that focused on carrying out Communist Party directives. McKinsey is also advising the Saudi crown prince, Mohammed bin Salman, as he seeks to make its economy less reliant on oil.

While confidentiality is necessary in private business, it can become problematic when public money is involved, as in South Africa, or for that matter in the United States, where McKinsey has advised more than 40 federal agencies, including the Federal Bureau of Investigation, the Central Intelligence Agency, the Defense Department and the Food and Drug Administration.

Since President Trump took office, McKinsey has greatly expanded consulting for Immigration and Customs Enforcement through that agency’s office of “detention, compliance and removals.” Their contracts with the agency exceed $20 million. Asked about those contracts, a McKinsey spokesman said the company’s work focused primarily on administration and organization and was unconnected to immigration policy, including the separation of children and parents at the border.

Certainly, consulting firms other than McKinsey keep client lists confidential and work for authoritarian governments. And McKinsey has undeniably been a force for good, through its pro bono work and by helping many organizations become more efficient engines of economic growth. As for the quality of people McKinsey hires, many have gone on to run some of the world’s biggest and most successful companies.

The deal McKinsey struck with South Africa’s struggling state-owned power company, Eskom, had a potential value of $700 million.CreditGulshan Khan for The New York Times

The Firm Rewrites the Rules

In 2012, a new class of recruits — the worker bees in McKinsey’s hive — settled in at its office in Sandton, Johannesburg’s financial center, often called the richest square mile in Africa. They were part of a notably diverse group. Given South Africa’s historic battle against apartheid, it was a point of pride at McKinsey that more than 60 percent of the office’s 250 employees were black South Africans. Many described their work as a calling, an opportunity to make a difference in a young and still-struggling democracy.

Years earlier, McKinsey’s South African partners had decided that to be relevant, they had to embrace the public sector, because of its outsize role in South Africa’s economy. But a South African government weakened by corruption also represented a risk to McKinsey’s sterling reputation — a reputation forged by its founder, a math whiz from the Ozarks named James O. McKinsey, and nurtured by his disciple and successor, Marvin Bower.

Over the decades, the firm — that’s what it calls itself — became confidant to chief executives and presidents, simultaneously the secret-keeper of corporate America and its most effective evangelist, preaching the McKinsey way to companies across the globe.

In 1952, McKinsey helped the incoming President Dwight D. Eisenhower staff his new administration. Later it helped to set up NASA, and helped to invent the Universal Product Code — the bar code. The company was instrumental in Wall Street’s rise as a dominant force in the economy, providing SWAT teams of brainpower to help Merrill Lynch, Citigroup and countless other financial companies adapt and move into new markets.

There came to be a not entirely hyperbolic narrative of McKinsey’s preordained white-shoe path through the world — the Harvard Business School recruit turned McKinsey consultant turned rising corporate titan. And few companies have a better track record of producing them: Louis V. Gerstner Jr., who oversaw I.B.M.’s turnaround in 1990s, is a McKinsey veteran. So is Sheryl Sandberg of Facebook, and Google’s chief executive, Sundar Pichai.

“So pervasive is the firm’s influence today that it is hard to imagine the place of business in the world without McKinsey,” wrote Duff McDonald, author of “The Firm,” a 2013 book about the company.

McKinsey has had its share of bad publicity, but much of it has focused on people who have already left.

One very public flap emerged in the summer of 1970, when The Times (also a McKinsey client over many decades) published front-page articles detailing an explosion in consulting fees paid out by New York City. At the center of the controversy was a young McKinsey partner, acting as an unpaid official in the city’s budget bureau even as the city was spending taxpayer dollars on contracts with the firm.

It looked bad. And while McKinsey was cleared of wrongdoing, the experience helped steer the company away from government work, avoiding the publicity, the ethical quandaries and the generally lower fees that came with public contracts.

But if McKinsey had learned a lesson, it soon began to unlearn it.

By the early 2000s, McKinsey re-entered the public sphere in a major way — and now government contracts and work with state-owned companies make up 16 percent of the firm’s revenues.

There was another lesson being unlearned as well: Work for a fixed fee.

In the late 1980s, an up-and-coming partner in McKinsey’s energy practice, Jeffrey K. Skilling, had been part of a committee considering whether payment should be based on delivered results, such as reduced costs.

As Mr. Skilling told the journalist Anita Raghavan, the panel concluded it would not work, because getting paid based on impact, for example, could give McKinsey an incentive to tell clients to reduce costs even if it was not in their interest. Doing that, Mr. Skilling said, “could destroy” the firm.

Mr. Skilling — who would become Enron’s chief executive and end up in federal prison after its vast accounting fraud was revealed — saw the ethical trap. A future generation of McKinsey partners came to a different conclusion.

Starting around 2001 or 2002, McKinsey again began to rethink its fee-for-service rule. Its competitors were already changing over. After years of deliberation and study, the firm agreed in 2011 to allow “at risk” contracts alongside its traditional fee structure.

“There’s been client demand for that, clients saying we like that approach,” Mr. Barton said. “If you don’t get the results you want, then don’t pay us.”

With this new pay policy and avid embrace of government work, the firm’s South African partners had been handed a seductive vision of the future. Some of McKinsey’s young associates in Johannesburg would end up on the ill-fated Eskom deal. But first, there was an ambitious project at the state-owned rail and port agency, Transnet.

Eskom’s Lethabo Power Station. It is not clear that the flailing power company is much better off now than it was before the McKinsey deal.CreditGulshan Khan for The New York Times

‘Trying to Play God’

McKinsey had worked with Transnet since 2005, embedding itself so deeply that one board member wondered how the agency could ever oust the consultants should the need arise. Still, Transnet remained an underachiever, its ports inadequate, its freight rail system moribund.

Then, in February 2011, Transnet got a new chief executive, Brian Molefe, who had been running the country’s public pension fund.

His tenure began with controversy. The South African media had already linked him to the Guptas, a family led by three brothers who arrived in South Africa a quarter-century ago and became ostentatiously wealthy through a web of businesses, once commandeering an air force base to fly in wedding guests from India.

McKinsey and Mr. Molefe set out to revitalize the agency by buying as many as 1,064 new locomotives in what would be the biggest government procurement in South African history. But McKinsey would have to take on a subcontractor, under a South African law requiring companies that worked with state-owned enterprises to have black-owned partners.

According to prosecutors, the Guptas saw these black-empowerment companies as a way to empower themselves, and state-owned companies like Transnet became willing accomplices. Transnet steered McKinsey toward working with a company, Regiments, owned in part by a businessman linked to the Guptas.

But weeks before the winning bidders were announced, McKinsey bowed out, saying the process was moving too quickly.

Before becoming chief executive of Eskom, Brian Molefe led the state-owned rail and port agency, Transnet, which had also worked with McKinsey. He had also been linked to the Guptas, an influential family with ties to former President Jacob Zuma.CreditWaldo Swiegers/Bloomberg, via Getty Images

As it turned out, Transet agreed to pay about $1 billion more than the agreed-upon price for the locomotives. And, as leaked documents published last year in the local media revealed, one of the winning bidders, a state-owned Chinese company, paid more than $100 million to shell companies tied to another Gupta associate, Salim Essa.

Although it is unclear what, if anything, Mr. Molefe knew about those payments, he left Transnet to search for a new challenge. He found it in 2015 as the chief executive at Eskom.

The power company had long been the public’s favorite punching bag, notorious for its high rates, sputtering from one crisis to the next. Officials worried about getting enough coal, about delaying maintenance to keep electricity flowing. During the World Cup in 2010, Eskom feared that the lights might go out at any moment with the whole world watching.

To address Eskom’s financial troubles, McKinsey and Eskom drew up an audacious new reorganization plan.

McKinsey’s team leader on the project was a popular partner, Vikas Sagar, a stylish, Porsche-driving fitness buff in his 40s, known for hugging colleagues when the spirit moved him and fiercely charting his own course. He was assisted by Alexander Weiss, a serious reverse image of Mr. Sagar, who thought little of commuting between his home in Germany and Johannesburg.

McKinsey’s proposal appeared perfect for a company in desperate financial straits. Eskom would pay only if the plan produced savings. Then the consultancy would get a percentage. All the risk, ostensibly, would be McKinsey’s, since it might spend heavily but get nothing in the end.

Yet for all the upside, the proposal had a Trojan-horse quality: Eskom would hire McKinsey not knowing what the final bill would be.

The plan left several McKinsey partners uneasy. Could Eskom absorb and apply McKinsey’s recommendations? And how would a contract with an anticipated payout in the hundreds of millions of dollars be received by South Africans? Also troubling was the fact that McKinsey had won the contract without competitive bidding.

“You are betting the office,” one former partner recalled warning colleagues. If the final payout became public, that official added, “You are going to be slaughtered just for the size.”

The contract’s structure — with the risks it posed for McKinsey — was not universally embraced, either. “Trying to do a 100 percent at-risk contract at Eskom is trying to play God,” a former partner said. “You are really guaranteeing that I can turn around everything, no problem.” To accomplish that, McKinsey might need more political clout and expertise than it could deliver.

 

Most of McKinsey’s current or former partners who spoke to The Times requested anonymity because they were not authorized to speak to the media. McKinsey did provide several partners for interviews on the condition that their names not be used. Mr. Sagar did not respond to repeated messages seeking an interview, and Mr. Weiss declined to speak to The Times.

Concerns notwithstanding, the prospect of a big payday made the contract popular not only in Johannesburg but throughout McKinsey’s global empire. Supporters included two senior partners with oversight in energy and power: Yermolai Solzhenitsyn, the novelist Aleksandr Solzhenitsyn’s eldest son, in Moscow, and Thomas Vahlenkamp in Düsseldorf, Germany. Both declined to be interviewed.

In the end, Mr. Sagar and his allies carried the day.

In situations like these, risk managers are supposed to serve as corporate lifeguards, ready to whistle back dealmakers if they expose the company to unnecessary legal and reputational peril. Yet the Eskom contract was approved with less scrutiny than regular public contracts. That was because state-owned enterprises were treated as private corporations, where reviews focused on commercial viability, not political risk.

Had McKinsey vetted the Eskom contract properly, it might have spared itself some of the grief to come. The contract, it turned out, was illegal: The power company had failed to get a government waiver from the standard fee-for-service payment, despite assuring McKinsey that it had done so.

“For the scale of the fee, they were prepared to throw caution to the wind, and maybe because they thought they couldn’t be touched,” said David Lewis, executive director of Corruption Watch, a local advocacy group.

A shop in Cape Town during a rolling blackout in 2015. Eskom used scheduled outages to try to prevent a grid collapse.CreditMike Hutchings/Reuters

A Mystery Partner Is Unmasked

If McKinsey fell short in vetting the Eskom contract, the same could be said about the scrutiny of its minority partner, a company called Trillian Management Consulting.

McKinsey’s putative marriage to Trillian produced its first awkward moments when its chief executive, Bianca Goodson, showed up angry at the consultancy’s Sandton headquarters on a January evening in 2016. Under McKinsey’s agreement with Eskom, Ms. Goodson’s company was supposed to provide consulting support, but with only two employees was unsure how to do that.

Feeling ignored and marginalized, Ms. Goodson planned to raise her concerns at a meeting of McKinsey partners. Four hours in, she got her chance. But before she could finish, Ms. Goodson wrote in an account later submitted to Parliament, a McKinsey team leader abruptly left the room, another partner said not to worry too much about work because she would still get her financial cut, and another made “whipping sounds and gestures,” an apparent inside joke, prompting laughter among the partners.

Ms. Goodson left more disillusioned than before she arrived. Two months later, she resigned.

During the internal debate over the Eskom deal, several partners had questioned whether McKinsey knew enough about who precisely was behind Trillian. Now rumors began reaching the McKinsey office that Trillian Management and its parent company, Trillian Capital, might have ties to the Gupta family.

McKinsey knew little about Trillian — a new company, with no track record, that had broken off from McKinsey’s previous minority partner, Regiments, after a business dispute. What’s more, Trillian had refused McKinsey’s requests to divulge its ownership.

The Eskom affair is now part of a government investigation into how the Gupta family used its connections to manipulate and control state-owned enterprises for personal gain.CreditThuli Dlamini/Gallo Images, via Getty Images

Even so, McKinsey chose to kick the can down the road and continue working. What McKinsey did not yet know was that Eskom’s chief executive, Mr. Molefe, had placed dozens of phone calls to one of the Gupta brothers during and after contract negotiations.

An influential senior partner in Johannesburg, David Fine, had grown increasingly uneasy about Trillian, according to his testimony to Parliament. One source of concern: Over the objections of two senior partners, McKinsey’s team leader, Mr. Sagar, had been meeting with Eskom and Trillian without any other McKinsey officials present.

 

Eventually McKinsey hired a private investigative firm to dig into Trillian’s background. When that did not produce any definitive leads, Mr. Fine began running internet searches on companies named Trillian and found the name “S. Essa” listed as a director. Weeks later, the South African media revealed the majority owner of Trillian as none other than Salim Essa, the Gupta associate whose shell companies had received more than $100 million in the locomotive deal.

On March 30, 2016, McKinsey told Eskom in writing that it was severing its ties to Trillian. But while McKinsey had finally taken a stand, it quietly undercut that decision by continuing to work alongside Trillian — independently, rather than as a subcontractor.

To the consternation of some McKinsey partners, that arrangement continued until the end of June 2016. With the local media revealing ever more of the Gupta family’s influence, Eskom — not McKinsey — prematurely terminated the contract. Mr. Molefe resigned that November. Mr. Molefe did not respond to requests for comment for this article; a lawyer for the Guptas declined to comment.

The abbreviated tab for barely eight months of work: nearly $100 million, with close to 40 percent going to Trillian.

In the United States, with an economy over 50 times as big as South Africa’s, a contract that size might have gone unnoticed. But in South Africa, millions of dollars flowing out of a struggling public utility and into the pockets of consultants driving Porsches and Ferraris created an unsavory image that required a response.

Yet McKinsey kept quiet, one of many decisions the firm would come to regret.

A Surge of Public Scrutiny

Picketers outside McKinsey’s Sandton offices in October protested the firm’s dealings with Eskom and a partner, Trillian Management Consulting, about which McKinsey knew little.CreditFelix Dlangamandla/Netwerk24, via Gallo Images

Late the next year, South Africa’s National Prosecuting Authority would deliver a stinging summation of the Eskom case. McKinsey, the prosecutors would allege, had been instrumental “in creating a veil of legitimacy to what was otherwise a nonexistent, unlawful arrangement.” That arrangement, in turn, allowed a company controlled by the Gupta associate, Mr. Essa, to profit.

 

That conclusion was based in part on a letter obtained by a widely respected human-rights advocate, Geoff Budlender, who had been asked to investigate Trillian, including its ties to McKinsey. For the first time, McKinsey was being publicly held to account.

Mr. Budlender asked to interview McKinsey but was told to put his questions in writing, which he did. In response to one question, McKinsey denied working “on any projects” with Trillian, as either a subcontractor or a black-empowerment partner.

With his trap laid, Mr. Budlender pounced. He attached a Feb. 9, 2016, letter from the McKinsey team leader, Mr. Sagar, to Eskom. “As you know,” Mr. Sagar had written, “McKinsey has subcontracted a portion of the services to be performed” to Trillian. The letter went further and authorized Eskom to pay Trillian directly, rather than through McKinsey, as was customary for a subcontractor.

Asked to explain the conflicting answers, a McKinsey lawyer, Benedict Phiri, took weeks to respond, saying he needed to speak with his colleagues. Finally, he wrote that, given ongoing legal disputes, it was “inappropriate” to comment.

Mr. Budlender concluded that McKinsey’s denial was false. “I have to say that I find this inexplicable, particularly having regard to the fact that McKinsey presents itself as an international leader in management consulting and given the widespread public interest in this matter,” he wrote.

In the interview with The Times, the McKinsey managing partner, Mr. Barton, said the office leadership in Johannesburg had been unaware of Mr. Sagar’s letter, and had only learned of it from Mr. Budlender. But three current or former McKinsey partners told The Times that Mr. Sagar’s German colleague, Mr. Weiss, and the firm’s lawyer, Mr. Phiri, also knew of the letter.

 

McKinsey’s lawyers said that the letter should never have been sent. Even so, they said, the authorization to pay Trillian referred to another, much smaller contract, and it was predicated on Trillian’s meeting certain conditions.

Image

Bianca Goodson, former chief executive of Trillian Management Consulting, became a whistle-blower.CreditCarte Blanche

In late 2017, a parliamentary committee began calling witnesses as part of its own investigation of state capture. One witness was Ms. Goodson, the former Trillian executive, who said she had been told soon after being hired that Mr. Essa owned Trillian. She also testified about meeting Mr. Sagar and Mr. Essa in Melrose Arch, a wealthy enclave with high-end retail and sidewalk dining where deals are made.

McKinsey sent Mr. Fine, who withstood nearly four hours of questioning. He addressed criticism head on, beginning with the size of the contract. “We should have absolutely had a fee structure that was capped,” he said.

Mr. Fine, who had no role in the Eskom contract, said he had been assured that Eskom did derive measurable benefits from McKinsey’s consulting. Yet his comments betrayed an element of doubt. As a native South African, he said, he couldn’t help asking himself, “If these benefits were there, why then has the price of electricity gone up and has the liquidity position of Eskom deteriorated?”

Mr. Barton, in his interview with The Times, insisted that his firm helped Eskom solve important problems. He expressed frustration at the overarching narrative that McKinsey took money for little work. “There was real work being done,” he said.

Grieve Chelwa, an economics lecturer at the University of Cape Town, said in an interview that McKinsey’s top ranks in South Africa were overwhelmingly filled by Europeans who “may not have had the political antennae” to pick up potential problems.

“The less charitable interpretation is that they knew,” said Dr. Chelwa, until recently a fellow at Harvard’s Center for African Studies. “They made a risk calculation that we know what is going on or we have an idea what is going on, but then there is 1.6 billion rand to make, and what is the probability that all this falls in our faces? They made that kind of calculation and they said, ‘O.K., the risk is worth doing,’ and they did it.”

The proceeds of the Eskom contract have been frozen, pending the government’s investigation.CreditGulshan Khan for The New York Times

Hard Lessons Learned

It is a risk McKinsey now regrets taking.

The advocacy group Corruption Watch referred the firm’s conduct to the United States Justice Department for possible violations of the Foreign Corrupt Practices Act. McKinsey declined to say whether federal investigators had contacted the firm; the Justice Department declined to comment. The National Prosecuting Authority in South Africa has frozen the proceeds of the Eskom contract, pending the completion of the government’s investigation. And several banks and corporations, including the South African arm of Coca-Cola, have said they will not do business with McKinsey until investigations are concluded.

McKinsey vehemently denies breaking any laws, and says that this view has been validated by a monthslong internal inquiry involving more than 50 lawyers reviewing millions of documents and emails.

The firm does admit mistakes. McKinsey will now give state-owned companies the same scrutiny it would government agencies or ministries. That policy may have a major impact in China, where McKinsey has advised at least 19 of the biggest state-owned companies as well as the country’s powerful planning agency.

In a statement, McKinsey said that it was “not careful enough about who we associated with,” that it should not have worked alongside Trillian after cutting its ties and that it did not communicate properly with Mr. Budlender. “We are embarrassed by these failings, and we apologize to the people of South Africa, our clients, our colleagues and our alumni, who rightly expect more of our firm.”

At the end of June, Mr. Barton, 55, will step down as previously planned. McKinsey’s nearly 600 senior partners voted to replace Mr. Barton with Kevin Sneader, a British citizen. Last weekend, as The Times was preparing this article after weeks of questioning McKinsey about its secretive culture, The Financial Times published an interview with Mr. Sneader, who said the firm could no longer “hide from the outside world.”

Mr. Sagar has left the firm with his full benefits in place. Mr. Weiss has been sanctioned, though McKinsey declined to say what that involved. The firm’s Johannesburg lawyer, Mr. Phiri, resigned, and the head of McKinsey’s Africa practice was transferred to Hong Kong.

Mr. Fine, who now leads McKinsey’s worldwide public-sector practice in London, cast the fallout from Eskom in personal terms. “I have seen the anger and disappointment in my clients’ eyes,” he told the South African Parliament. He added, “I’ve experienced rejection from people that I really love and trust, and that’s been hard.

Walt Bogdanich reported from Johannesburg, and Michael Forsythe from New York. Reporting was contributed by Agustin Armendariz, Grace Ashford, Audrey Jiajia Li, Caterina Barbera, Eileen Grench, Bridget Hickey, Natasha Rodriguez and Anneke Ball Schnell in New York.

A version of this article appears in print on , on Page A1 of the New York edition with the headline: How a Top Adviser Blundered Into a Corrupt Deal.

Continue Reading

Advertising

GOOGLE MAKES $550M STRATEGIC INVESTMENT IN CHINESE E-COMMERCE FIRM JD.COM

Published

on

By

Google has been increasing its presence in China in recent times, and today it has continued that push by agreeing to a strategic partnership with e-commerce firm JD.com, which will see Google purchase $550 million worth of shares in the Chinese firm.

Google has made investments in China, released products there and opened offices that include an AI hub, but now it is working with JD.com largely outside of China. In a joint release, the companies said they would “collaborate on a range of strategic initiatives, including joint development of retail solutions” in Europe, the U.S. and Southeast Asia.

The goal here is to merge JD.com’s experience and technology in supply chain and logistics — in China, it has opened warehouses that use robots rather than workers — with Google’s customer reach, data and marketing to produce new kinds of online retail.

Initially, that will see the duo team up to offer JD.com products for sale on the Google Shopping platform across the word, but it seems clear that the companies have other collaborations in mind for the future.

JD.com is valued at around $60 billion, based on its NASDAQ share price, and the company has partnerships with the likes of Walmart and it has invested heavily in automated warehouse technology, drones and other “next-generation” retail and logistics.

The move for a distribution platform like Google to back a service provider like JD.com is interesting since the company, through search and advertising, has relationships with a range of e-commerce firms, including JD.com’s arch rival Alibaba.

But it is a sign of the times for Google, which has already developed relationships with JD.com and its biggest backer Tencent, the $500 billion Chinese internet giant. All three companies have backed Go-Jek, the ride-hailing challenger in Southeast Asia, while Tencent and Google previously inked a patent-sharing partnership and have co-invested in startups such as Chinese AI startup XtalPi.

 

 

Source: Tech Crunch.

Continue Reading

Industry

GOOGLE LAUNCHES A PODCAST APP FOR ANDROID WITH PERSONALIZED RECOMMENDATIONS

Published

on

By

Google today is introducing a new standalone podcast app for Android. The app, called simply Google Podcasts, will use Google’s recommendation algorithms in an effort to connect people with shows they might enjoy based on their listening habits. While podcasts have previously been available on Android through Google Play Music and third-party apps, Google says the company expects Podcasts to bring the form to hundreds of millions of new listeners around the world. (Google Listen, an early effort to build what was then called a “podcatcher” for Android, was killed off in 2012.)

“There’s still tons of room for growth when it comes to podcast listening,” said Zack Reneau-Wedeen, product manager on the app. Creating a native first-party Android app for podcasts “could as much as double worldwide listenership of podcasts overall,” he said.

Google Podcasts will look familiar to anyone who has used a podcast app before. It lets you search for new podcasts, download them, and play them at your convenience. More than 2 million podcasts will be available on the app on launch day, Google says, including “all of the ones you’ve heard of.”

Open the app, and a section called “For you” shows you new episodes of shows you’ve subscribed to, episodes you’ve been listening to but haven’t finished, and a list of your downloaded episodes. Scroll down, and you’ll see top and trending podcasts, both in general and by category. The podcast player has fewer fine-grained controls than you might be used to from apps like Overcast. You can’t customize the skip buttons or create playlists of podcasts to listen to, for example.

The Podcasts app is integrated with Google Assistant, meaning you can search for and play podcasts wherever you have Assistant enabled. The company will sync your place in a podcast across all Google products, so if you listen to half a podcast on your way home from work, you can resume it on your Google Home once you’re back at the house.

In the coming months, Google plans to add a suite of features to Podcasts that are powered by artificial intelligence. One feature will add closed captions to your podcast, so you can read along as you listen. It’s a feature that could be useful to people who are hard of hearing or for anyone who is listening in a noisy environment. (I usually miss a few minutes of the podcasts I listen to every day, thanks to a noisy subway ride.)

Closed captions also mean that you’ll be able to skip ahead to see what’s coming up later in a show. Eventually, you’ll be able to read real-time live transcriptions in the language of your choice, letting you “listen” to a podcast even if you don’t speak the same tongue as the host.

Google also wants to expand the number of people making podcasts. The company’s research showed that only one-quarter of podcast hosts are female, and even fewer are people of color. In an effort to diversify the field, Google formed an independent advisory board that will consider ways to promote podcast production outside of the handful of major metropolitan areas in the United States that currently dominate the field.

Google will not pay any creators to make podcasts directly, the company said, but it will likely explore ways of giving podcasts from underrepresented creators extra promotion. It’s also examining ways to make recording equipment more accessible to people who can’t afford it.

If you already listen to podcasts on Google Play Music, nothing will change today. But the company made it clear that it plans to focus its future efforts around podcasting in the standalone app.

The Android app can be downloaded here. There are currently no plans for an iOS app.

Continue Reading

Subscribe to Blog via Email

Enter your email address to subscribe to this blog and receive notifications of new posts by email.

Join 680 other subscribers

Advertisement

Trending

    %d bloggers like this: