Microsoft Just Cut 3,000 More Jobs — 8 Charts on What Headcount Changes Really Mean for Big Companies

One person losing a job is a tragedy. Ten thousand people losing their jobs is…?

Jerry Berg worked in Building 26 at Microsoft's Redmond headquarters. A senior software test engineer, he had been at Microsoft for 15 years, writing automated test programs that put machines to work hunting for bugs in the Windows operating system.

One day in July, he was suddenly called into a room with about 150 colleagues, most of them working on Windows. Nobody had any idea what the meeting was about.

An executive soon walked to the front and said Microsoft was changing direction and their jobs were no longer needed, then brought up HR to explain the severance terms. Roughly: everyone would get some compensation, and everyone had four days to hand over their work and turn in their badges.

Later, recalling the day, Berg remarked dryly that this was probably the best-kept secret in Microsoft's history.

It happened on July 17, 2014, when Microsoft announced the largest layoffs in its history — 18,000 people. Because the cuts mainly targeted Nokia employees who had joined Microsoft only months earlier, ordinary Windows staff like Berg were caught off guard. Microsoft's executives were probably less surprised — buying Nokia's phone business had been meant to give the Windows division a new revenue stream from hardware.

Three years later, almost the same thing happened again. On July 7 this year, Microsoft announced it would cut 3,000 to 4,000 positions, mostly in sales. About 10% of Microsoft's entire sales force would go.

Microsoft is now in its 42nd year. Its first big layoffs came in 2009, and since then the cuts have come faster and taken more people. Layoffs are no longer news at Microsoft: for three years running, thousands of people have lost their jobs every July.

The reason given is usually the one Berg heard — the company is changing direction and no longer needs certain people.

Looking back on his years at Microsoft, former CEO Steve Ballmer has said his biggest regret was failing to line up hardware talent early enough — failing to turn Microsoft into a hardware company, failing to have people in place who could beat Apple.

The reverse holds too. When Satya Nadella, Ballmer's successor, decided in 2014 to abandon phones, the people Microsoft had bought from Nokia were of no use to it.

The 3,000 people cut this time went for a similar reason. Microsoft now cares more about selling businesses subscription-based online Office services and cloud computing than about selling Windows and Office licenses by the copy. The new work calls for salespeople with different experience.

Looking back, every round of layoffs since 2009 has been the price Microsoft paid for a failed search for its future. The damage reached from executives — CEOs, CTOs, executive vice presidents — down to tens of thousands of rank-and-file employees and even interns.

Every Layoff Is the Price Microsoft Pays for Searching for Its Future

A large share of Microsoft's layoffs are the cleanup after one particular business failed. The Zune music player, which went up against Apple's iPod, is one example; Bing Maps, abandoned two years ago, is another.

1. 2009: Layoffs in Microsoft's Zune Division

• Time: February 2009
• Reason: the Zune business failed
• Scale: Microsoft acknowledged layoffs in the Zune division but did not disclose numbers
• Failed business: music players

Strategic shift at the time: the Zune software and services team was folded into the Connected TV division (Windows Media Center, Mediaroom), and the Zune hardware team into the mobile communications business.

Aftermath: in early 2011, Ballmer announced Zune was being discontinued.

2. 2015: Microsoft Sells Off Bing Maps

• Time: June 2015
• Reason: Bing Maps was sold to Uber
• Scale: 100 people
• Failed business: Bing had lost money for years, and even in the U.S., where Microsoft had spent heavily to buy share, it held only about 10% of the market. Maps mainly served phones, and Microsoft's phone business had failed badly.

Strategic shift at the time: phones were abandoned altogether, and Bing turned to cutting its losses.

Most of these failures were products that sold badly after launch and that Microsoft then dropped. The advertising business was cut back several times for the same reason. But some projects were killed outright after losing internal fights — the Courier tablet, incubated inside the Xbox division, is one; its failure sent several Xbox executives out the door.

After the iPhone's rise, Microsoft went through two big structural overhauls. Both CEOs tried to reorganize their way to a new future, and both reorganizations came with layoffs of thousands of people.

3. 2009: Major Microsoft Layoffs

• Time: January 2009
• Failed business: Vista, plus company-wide results hit by the financial crisis (Microsoft's second-quarter profit was $4.71 billion, down 11% year over year)
• Scale: 5,800 people. Deep cuts hit sales, marketing and services, IT, Zune, gaming advertising, and other units.
• Executive departures: Ray Ozzie, then Microsoft's chief software architect, left. Bill Gates himself had held that title before Ozzie.

Strategic shift at the time: Ballmer reorganized Microsoft, splitting the three original business-area divisions into five independent business units, each with its own marketing team.

After the reorganization, every Microsoft business unit answered for its own results. The new structure gave Office and cloud computing better room to grow. But it also killed collaboration across divisions, making it hard to build products that combined hardware and software — Xbox was the one exception — and leaving Microsoft's phone business with no way to fight back against Apple and Google.

4. 2014: Major Microsoft Layoffs

• Time: July 2014
• Failed business: phones and the Surface tablet
• Scale: Nadella cut 18,000 jobs, 70% of them from the former Nokia phone division. Over the next two years, Microsoft carried out several more rounds of layoffs of over a thousand people each, all aimed at the phone business.
• Executive departures: Ballmer retired early; Stephen Elop, the former Nokia CEO who ran Microsoft's phone business, was left without a division of his own and departed soon after.

Strategic shift at the time: dropping the principle that everything revolves around Windows.

The 2014 layoffs marked the biggest transformation in Microsoft's history, and it came in fits and starts. It began in 2013, when Ballmer bought Nokia's phone business and soon after announced the "One Microsoft" reorganization, trying to turn Microsoft from a set of separate business units into an Apple-style structure in which divisions worked together, so it could fight Apple on hardware.

But Ballmer's plan met resistance from a Microsoft board that included Gates, and he ended up announcing his early retirement before the company had found a successor.

In February 2014, Nadella took over as CEO. He first cut most of the former Nokia employees, then merged the hardware business into Windows. Forced to carry the hardware losses, the Windows division gave up on phones entirely a year later. Two years on, Nadella threw out the hardware business Ballmer had started and brought Office and cloud computing to Android and iOS. The whole company's focus became selling enterprise services on every platform, rather than revolving around Windows.

Yet after all these rounds of layoffs, Microsoft has not ended up with fewer people. Leave aside the more than 30,000 employees who arrived all at once when Microsoft bought Nokia in 2013, and Microsoft's headcount has in fact grown steadily.

Ballmer said more than once while running Microsoft that starting a new business is a gamble, and that if you are going to gamble, you should gamble big. If all those rounds of layoffs were the price of the losing bets, then the enterprise sales machine he built and the cloud business that followed were the winning ones.

Microsoft's latest 2017 figures show net profit of $4.8 billion last quarter. For comparison, Google's revenue in the first quarter of 2017 was $24.75 billion with profit of $5.43 billion, and Alibaba's revenue was $5.605 billion with profit of $1.431 billion.

Phones never worked out, and the Surface tablet has always been a niche product. But Office, Windows, and cloud services sold to businesses keep bringing Microsoft real money. Alongside this round of 3,000 layoffs, Microsoft also announced a hiring plan: it will take on more people to sell cloud services.

When a company no longer has a core business it can count on, though, layoffs get a lot uglier. Motorola, for instance.

Without a Business to Rely On, There Are No Employees Left Either

In September 2016, Lenovo announced global layoffs of 3,200 people, including 500 at Motorola. After a run of spinoffs, acquisitions, and sales, the Motorola phone business — the business that first put mobile phones in consumers' hands — was down to 700 people.

Motorola's decline started with Nokia's rise in the 1990s. After Android arrived, Motorola picked up briefly. Verizon, the largest U.S. carrier, worked with Motorola on the Droid (sold in China as the Milestone) to take on the iPhone, which the second-largest carrier, AT&T, carried. But in the global market Motorola had no answer to Apple, Samsung, and HTC.

In 2011, after years of losses, Motorola split into two companies: Motorola Mobility, which made phones, and Motorola Solutions, which made communications equipment. That same year, right after the split, Google announced it would buy Motorola Mobility for $12.5 billion.

Google then spent a while trying to build a hardware business around it, sending in executives and reworking the product line.

In 2013 the new Moto X arrived, a genuinely inventive phone. It listened all the time, so you could wake the voice assistant without pressing a button. It also lit up the screen when you picked it up. Other phone makers, Apple included, later adopted both ideas.

Google also had the Moto X assembled in the United States and sold it on the slogan "the only smartphone assembled in the USA."

Beyond the phone business itself, Google bought Motorola with an eye on the patent war then raging between Android and Apple, and on keeping Samsung in check. Samsung had been going further and further with its own phone software: it had announced plans for the Tizen operating system, and it was customizing the interface on its Android phones more and more.

But the Moto X did not sell. Three months after launch it had sold 500,000 units worldwide. Meanwhile Google and Samsung struck a deal: Samsung shared patents with Google and gave up further deep customization of the system. Tizen was later repurposed for watches and home appliances rather than used to challenge Android.

By then Motorola meant little to Google. In February 2014, Google sold it to Lenovo for a quarter of what it had paid. By the time of the deal, Google had already cut Motorola down to 3,800 people.

Lenovo already had several phone brands of its own, but none of them had become a real "brand" — most of its phones still sold through carrier channels.

With Motorola in hand, Lenovo's first thought was much like Google's: use the name to make mid- and high-end phones. In 2015, Lenovo brought the second-generation Moto X to China. It went on sale later than it had overseas and cost more than 3,000 yuan, and it barely made a ripple.

In 2016, Lenovo launched the modular Moto Z, which let users clip a bigger camera module, a speaker, or a battery onto the back. The Moto Z did not turn Lenovo's losses around.

In the end Lenovo began a new round of layoffs that hit several phone brands, cutting Motorola's staff to 700. The Motorola name has not been abandoned — under Lenovo's new plan, the letters "Moto" will actually appear on more new products.

It is just that those letters are all that is left of the Motorola that first brought mobile phones to market.

More People Isn't Always Better, But a Company's Headcount Usually Tracks Its Fortunes

Walmart is the world's largest employer, with 2.3 million employees worldwide; in its home market, the United States, 1% of the working population works for Walmart.

By 2015 this giant had grown to 11,767 stores worldwide and $485.65 billion in retail sales. But the rival it is always compared with, Amazon — founded more than 30 years after Walmart — had passed it.

Counted by employees, Amazon had only 230,000 people in 2015, roughly a tenth of Walmart's total.

One company opens stores everywhere and expects customers to drive to a big-box store and carry their purchases home. The other has only a handful of experimental storefronts and expects customers to order online and wait for a courier at the door.

The two companies need people in different ways.

But look at how their headcounts grow, and it is not hard to tell which one is on the way up.

By the end of 2016, Amazon employed 341,000 people, 110,000 more than during the shopping season a year earlier — growth of nearly 50%. Most of the new hires work in Amazon's logistics network; in that one year Amazon built eight new fulfillment centers.

Over the past five years, Amazon has added 310,000 employees. Under a plan announced this January, it will add another 100,000 by the middle of next year, bringing the total close to 450,000.

Walmart's headcount has barely moved, holding between 2.1 and 2.3 million for years. Over the past decade one company's curve is flat and the other's climbs steeply — a close match to how the two businesses have grown over the past five years.

The same thing has played out in China. Baidu, Alibaba, and Tencent, once running neck and neck, began to grow at different speeds in 2014. Baidu's headcount spiked that year, just as it was pouring money into O2O, but the spike produced no breakthrough.

And these past few years have indeed been hard on Baidu.

More than 90% of Baidu's revenue used to come from advertising. After the company's advertising problems came to a head in 2016 and regulators stepped in, revenue stalled and profit fell sharply. The advertising business began shrinking in the third quarter of 2016, the first decline since Baidu went public in 2005. Its headcount is also the only one among the three giants to have fallen rather than risen.

It's Not Just Total Headcount That Matters to a Company — It's How Many People Leave

There is another number, easy to overlook, that matters when you weigh what a company's staffing changes mean: the turnover rate.

Set Amazon aside, since it fights mainly online. In physical retail, Walmart has always been readier than its peers to pay the lowest wages and hire the cheapest labor, and for years it has been criticized for low pay, excessive hours, gender discrimination, and suppressing union drives.

In 2015, Walmart set aside $1 billion to adjust employee pay. Behind that move was a staff turnover rate that had stayed stubbornly high for years. Figures compiled by Bloomberg show Walmart's turnover in 2012 at 21%, against 6% at another American brick-and-mortar retailer, Costco.

Behind the gap are two very different approaches to pay and benefits. In 2012, Costco paid $16 an hour; Walmart paid $11.50. On retirement benefits, health plans, and much else, Costco's package was far better.

Paying more to keep better people is part of Costco's strategy as a retailer. In 2012, each Costco employee generated $13,647 in profit, against $11,039 for each Walmart employee. On the face of it, better benefits cost Costco more.

But measured against sales, Costco's labor costs came to 9.8% of the total, against 17% at Walmart. That is because, once you count the full cost of employing people, turnover brings enormous hiring and exit costs.

To hold its headcount steady, Walmart has to hire 2.66 million store employees a year. Costco, with only 170,000 employees in all, needs to hire 180,000 a year to stay level.

The same turnover problem is working on the biggest company in the sharing economy, Uber. A JPMorgan study found that monthly earnings on labor platforms like Uber peaked in June 2014; once the subsidies ended, drivers' earnings started to fall, and a large share of the financially stable ones may have left this kind of work for good.

In early 2015, the labor economist Alan Krueger wrote in a paper that only 55% of Uber drivers were still active a year after signing up.

It got worse after that. By 2016, more than 75% of drivers who had completed at least one Uber trip left within a year.

Drivers who do not stay mean Uber has to keep recruiting new ones, and signing up a driver costs far more than keeping one.

A big reason so many left is that Uber kept cutting fares while taking a larger cut.

After leaving the Chinese market in 2016, Uber's losses eased somewhat, but they still ran to more than 60% of revenue. Uber's original story was that once it won the market it would cut driver subsidies and turn a profit. It now looks as though when the subsidies were cut, most drivers simply stopped driving.

Scandal followed scandal at Uber this year, and the investors finally joined forces to push CEO Travis Kalanick out. The scandals were only the fuse. Every underlying problem had been there for a long time; it was when Uber's growth story stopped holding up that the conflicts all caught fire. The CEO, the president, the CFO, and the head of engineering have now all left, which leaves the whole company running more or less driverless.

Technology Lets Each Person Do More, But Today's Biggest Companies Are Bigger Than They Were 50 Years Ago

When we talk about a company's headcount today, the number is routinely in the thousands or tens of thousands.

Before 1860, that would have been hard to imagine. In those days America's richest man, John Jacob Astor, ran the American Fur Company, and even at the peak of his career it employed fewer than 10 people.

That changed after 1860. The United States needed about 31,000 miles of railroad (by 1910, total track reached 240,000 miles), which took enormous capital and manpower. Once the lines were built, moving freight across the country and keeping trains from colliding took a huge administrative apparatus. That is how the first large corporations were born: by 1891 the entire U.S. Army numbered fewer than 40,000, while the Pennsylvania Railroad already employed more than 110,000.

Then, helped by the railroads and by the spread of electric power in factories, retail and manufacturing grew as well.

By 1960 the world's nine most valuable companies still came mostly from manufacturing, telecommunications, and petrochemicals, though technology and photography companies such as IBM and Kodak had begun to appear.

Today, with better technology and with basic work largely outsourced, many of those old giants are smaller. ExxonMobil, the oil company still among the ten most valuable, had 146,000 employees in 1960 and has 71,000 now. Petrochemicals, manufacturing, and the auto industry have all seen employment hit hard by technology. The National Bureau of Economic Research, the largest economic research institute in the United States, studied the labor market from 1990 to 2007 and found that every additional robot in an American factory wiped out 3 to 5.6 local jobs on average.

What is interesting is that if you compare the nine most valuable companies, today's giants do not employ fewer people.

Compared with 1960, today's nine most valuable companies employ 6% more people between them. But six of today's top nine are technology companies, and apart from Apple, nearly all of their contact with users happens through a screen.

More and more labor-intensive work has moved to the back end, where most people never see it. To most people, Google and Baidu are just a search engine. Amazon, Alibaba, and JD.com are just an app and a courier knocking at the door.

Amazon plans to grow to 500,000 employees over the next five years, mostly to build and maintain its logistics and warehousing. Facebook, the technology company with the fewest employees of them all, still has 17,048. This social network, with 2 billion users worldwide, already puts 4,500 people on content moderation alone, and after months of suicides livestreamed on the platform, Mark Zuckerberg announced in early May this year that Facebook would hire more than 3,000 additional moderators.

But when these companies run into the kind of trouble with a new business that Microsoft has faced, they do much the same thing. Google (Alphabet), in fact, already shows signs of it.

For a company, these swings in headcount may buy longer-term growth. Nadella's rounds of layoffs appear to have given Microsoft a concrete purpose again, and the stock has climbed steadily.

For each person caught up in them, the feeling is far more complicated. When Jerry Berg, the software test engineer with 15 years at Microsoft, got the news, the first thing he thought of was the cost of treatment for his autistic son.

"I kind of feel like I was laid off by a computer. A computer put me on the list for some reason or other." That is the layoff Berg remembers.

Cover image from Pinterest