ORGANIZATIONAL STRUCTURE

The End of Bureaucracy

by Gary Hamel and Michele Zanini
FROM THE NOVEMBER–DECEMBER 2018 ISSUE

B ureaucracy has few fans. Walmart CEO Doug McMillon calls it “a villain.” Berkshire
Hathaway vice chair Charlie Munger says its tentacles should be treated like “the cancers
they so much resemble.” Jamie Dimon, the CEO of JPMorgan Chase, agrees that
bureaucracy is “a disease.” These leaders understand that bureaucracy saps initiative, inhibits risk
taking, and crushes creativity. It’s a tax on human achievement.

Though mindful of its evils, many people believe bureaucracy is unavoidable. Dimon remembers an
outside adviser who defended it as the “necessary outcome of complex businesses operating in
complex international and regulatory environments.” Indeed, since 1983 the number of managers,
supervisors, and administrators in the U.S. workforce has grown by more than 100%, while the
number of people in all other occupations has increased by just 44%. In a survey by Harvard
Business Review, nearly two-thirds of respondents said their organizations had become more

/
bureaucratic in recent years. Peter Drucker’s prediction that today’s organizations would have half as
many layers and one-third as many managers as their late-1980s counterparts was woefully off the
mark. Bureaucracy has been thriving.

Meanwhile, productivity growth has stalled. From 1948 to 2004, U.S. labor productivity among
nonfinancial firms grew by an annual average of 2.5%. Since then its growth has averaged just 1.1%.
That’s no coincidence: Bureaucracy is particularly virulent in large companies, which have come to
dominate the U.S. economy. More than a third of the U.S. labor force now works in firms with more
than 5,000 employees—where those on the front lines are buried under eight levels of management,
on average.

Some look to start-ups as an antidote. But although firms such as Uber, Airbnb, Farfetch, and Didi
Chuxing get a lot of press, these and other unicorns account for a small fraction of their respective
economies. And as entrepreneurial ventures scale up, they fall victim to bureaucracy themselves.
One fast-growing IT vendor managed to accumulate 600 vice presidents on its way to reaching $4
billion in annual sales.

Why is bureaucracy so resistant to efforts to kill it? In part because it works, at least to a degree. With
its clear lines of authority, specialized units, and standardized tasks, bureaucracy facilitates
efficiency at scale. It’s also comfortably familiar, varying little across industries, cultures, and
political systems.

Despite this, bureaucracy is not inevitable. Since the term was coined, roughly two centuries ago,
much has changed. Today’s employees are skilled, not illiterate; competitive advantage comes from
innovation, not sheer size; communication is instantaneous, not tortuous; and the pace of change is
hypersonic, not glacial.

These new realities are at last producing alternatives to bureaucracy. Perhaps the most promising
model can be found at a company that would not, at first glance, appear to be a child of the digital
age. Haier, based in Qingdao, China, is currently the world’s largest appliance maker. With revenue
of $35 billion, it competes with household names such as Whirlpool, LG, and Electrolux. At present,
Haier has some 75,000 employees globally. Outside China it has 27,000 employees, many of whom
joined the company when it bought GE’s appliance business, in 2016.
/
Over the past decade the gross profits of Haier’s core appliance business have grown by 23% a year,
while revenue has increased by 18% annually. The company has also created more than $2 billion in
market value from new ventures. Those feats are unmatched by any of Haier’s domestic or global
competitors. This remarkable journey hasn’t been entirely pain-free. In recent years Haier dismissed
more than 10,000 employees. Yet it has also generated tens of thousands of new jobs in its rapidly
expanding ecosystem. Haier’s logistics network, which stretches across China, now includes more
than 90,000 independent drivers, for instance.

Microenterprises are free to form and evolve

with little central direction.
Haier’s success is the result of a root-and-branch overhaul of its once-traditional management
model. Having long viewed bureaucracy as a competitive liability, Zhang Ruimin, Haier’s renegade
CEO, has for a decade led an effort to build a company where everyone is directly accountable to
customers (a policy he describes as “zero distance”), employees are energetic entrepreneurs, and an
open ecosystem of users, inventors, and partners replaces formal hierarchy.

Haier’s shorthand for these practices is rendanheyi, a mash-up of Chinese characters that connotes a
tight coupling of the value created for customers with the value received by employees. The
rendanheyi model departs from bureaucratic norms in seven critical ways, which we’ll look at in
depth in this article.

1. From Monolithic Businesses to Microenterprises

Large corporations often consist of a few dominant businesses, each with its own orthodoxy about
strategy, customers, and technology. These tightly integrated entities and their monocultures make
a company vulnerable to unconventional competitors and blind it to new kinds of opportunities. To
avoid that risk, Haier has divided itself into more than 4,000 microenterprises, or MEs, most of
which have 10 to 15 employees. To be sure, some MEs, particularly in manufacturing, have larger
payrolls, but even in them decisions are made by small autonomous teams.

Microenterprises come in three varieties. First, there are roughly 200 “transforming” MEs—market-
facing units that have roots in Haier’s legacy appliance business but are reinventing themselves for
today’s customer-centric, web-enabled world. Zhisheng, which makes refrigerators for young urban
/
customers, is a typical example.

Second, there are 50-plus “incubating” MEs, or entirely new businesses. Some, like Thunderobot,
are focused on emerging markets such as e-gaming, while others, like Xinchu—a “smart” refrigerator
that connects users with third-party services that sell fresh food and deliver it within 30 minutes—
are wrapping new business models around familiar products.

Finally, there are roughly 3,800 “node” MEs. These businesses sell component products and
services such as design, manufacturing, and human resource support to Haier’s market-facing MEs.

Microenterprises are key to Zhang’s goal of building the world’s first company for the internet age.
That entails more than developing web-enabled products. It means creating an organizational model
that mimics the architecture of the internet: “small pieces, loosely joined,” as the Harvard
technologist David Weinberger famously put it. The web is incredibly diverse and yet still coherent.
While it has spawned countless innovations, it’s held together by common technical standards that
make cyberspace navigable and allow sites to swap resources like data.

Haier’s modular structure is similarly flexible but coherent. MEs are free to form and evolve with
little central direction, but they all share the same approach to target setting, internal contracting,
and cross-unit coordination.

2. From Incremental Goals to Leading Targets

In most organizations there’s little that supports bold thinking and doing. Old assumptions get
challenged only once the business has hit a wall. Not so at Haier.

Every ME is charged with pursuing ambitious goals for growth and transformation—known
internally as “leading targets.” Rather than taking last year’s performance as a starting point,
objectives are set “outside in.” A dedicated research unit collects product-by-product statistics on
market growth rates around the world and then uses that data to establish targets for the MEs.

Market-facing MEs are expected to grow revenue and profit four to 10 times faster than the industry
average. In product categories where Haier lags, the bar is set the highest, since there’s plenty of
room to increase share. In areas where Haier leads, the target is more modest but still a multiple of
/
the market baseline. As one ME’s leader put it, “If you’re number three or four in the market, leading
targets force you to think about how to become number one, and if you’re number one, they force
you to think about how to extend your lead.”

While ambitious, the targets do get adjusted when circumstances change. For example, when the
Chinese government revamped its energy efficiency standards for appliances in a way that favored
Haier’s already efficient refrigerators, the Zhisheng ME raised its targets.

Every market-facing ME is also expected to make a transformative leap from selling products and
services to building an ecosystem. A good example is Community Laundry. Having developed a
popular smartphone app that allows university students across China to schedule and pay for the
use of dormitory laundry facilities, the ME gave outside vendors access to the app’s more than 9
million users. Today the Community Laundry platform hosts dozens of other businesses and takes a
share of the revenue they generate.

Haier tracks the transformation of every ME with a “win-win value-added” statement, which
captures detailed metrics such as the extent of user involvement in product development, the
degree to which Haier’s products offer unique customer value, and the percentage of profits derived
from ecosystem revenue.

Every market-facing ME is expected to

eventually build a business ecosystem.
Like their market-focused siblings, node MEs have leading targets that are pegged to external
benchmarks. A manufacturing node, for example, may be responsible for lowering costs, cutting
delivery time, improving quality, and further automating its production facilities.

3. From Internal Monopolies to Internal Contracting

In most organizations, a significant percentage of employees are insulated from market forces. They
work in functions that are, in essence, internal monopolies, such as human resources, research and
development, manufacturing, finance, information technology, and legal affairs. However inept or
inefficient these internal providers may be, they can’t be fired by the units they serve.

/
At Haier every ME is free to buy services, or not, from other MEs. (A typical user ME will have
agreements with dozens of nodes.) If an ME believes that an external provider would better meet its
needs, it can go outside for services. Senior executives virtually never interfere with internal
negotiations.

Each year every market-focused ME looks at its performance objectives and asks itself, “What sort of
design, technology, production, and marketing support will we need to meet these goals?” Once it
has the answers, it asks the nodes for bids. Usually, two or three nodes will respond with proposals.
The ensuing discussions provide an opportunity for all parties to challenge existing practices and
brainstorm new approaches.

While the process may sound cumbersome, it’s facilitated by “presets”: predefined rules about
margin split and minimum performance standards that reduce friction during negotiations. Terms
can be renegotiated over the course of a year as circumstances change—hence Haier’s preference for
the word “agreement” over “contract.” One ME leader told us that he had replaced a dozen nodes
with other suppliers in the past 18 months. Nodes that are unable to provide competitive service can
and do go out of business.

A substantial part of a node’s revenue depends on the success of its ME customers. When a customer
unit fails to meet its leading targets, the node takes a hit. Every node is thus invested in the
performance of the market-facing units, and every employee’s pay is linked to market outcomes.
Zhang is only slightly exaggerating when he says, “At Haier we are no longer paying our employees.
Instead, they are paid by customers.”

This compensation model has three benefits. First, it discourages mediocrity. Nodes that don’t
deliver high levels of service lose their internal customers. Second, it unites everyone around the
goal of creating great customer experiences. When a user ME seems in danger of missing its targets,
representatives of all its supplier nodes quickly come together to resolve the problem. Third, it
maximizes flexibility: Market-facing MEs are free to reconfigure their network of service providers
as new opportunities emerge.

4. From Top-Down Coordination to Voluntary Collaboration

/
How does a company with more than 4,000 independent operating units synchronize major
investments in technology and facilities? How does it build cross-business capabilities such as
manufacturing automation?

New offerings don’t get a significant budget

until they’re validated by users.
In a start-up, coordination happens spontaneously. When there’s a problem people simply huddle
and hash things out. As a company grows and operating units become more siloed, coordination
becomes increasingly difficult. The typical solution involves more layers, mandates, and corporate-
level functions.

Haier has a different approach: organizing all MEs into platforms. Some platforms bring together
MEs operating in a similar category, like washing or audiovisual products, while others focus on
building new capabilities, such as digital marketing and mass customization. A typical industry
platform encompasses more than 50 MEs. (See the exhibit “A New Organizing Principle” for a
snapshot of one.)

A New Organizing Principle

Haier is made up of thousands of microenterprises (MEs), which are grouped into platforms. Below
is a map of the refrigeration platform.

/
 Any user ME is free to hire and fire nodes as it sees fit—or to go outside for services if it believes an
external provider can better meet its needs.

It’s the job of the platform owner to get ME teams together and help them identify opportunities for
collaboration, such as developing expertise in the internet of things. Critically, no one reports to the
platform owner, nor does the platform owner have a staff. Here’s how Wu Yong, the refrigeration
platform owner, describes his role in getting MEs to adopt a new frost-free technology, a move that
required an expensive upgrade of production facilities: “I helped facilitate, but the microenterprise
teams planned and executed the job together.”

Platform owners have leading targets and are expected to grow their platforms by developing new
MEs. In 2014, for instance, motivated by Haier’s goal of becoming the world leader in smart
appliances, Wu funded the networked refrigerator start-up Xinchu. Platform owners are as much
entrepreneurs as facilitators.
/
“Integration” nodes, found within every industry platform, help MEs import technology from across
Haier and identify internal partners that can coinvest in new initiatives. Like platform owners,
integration nodes encourage collaboration rather than exert control.

MEs also rely on the expertise of competence-focused platforms. Two of the most important are
smart manufacturing and marketing, each of which employs fewer than 100 individuals. The largest
node within the manufacturing platform provides technical support for mass customization.
Another node, smart engineering, deploys advanced production tools for the company.

The primary role of the marketing platform is supplying customer information. While every user ME
collects copious amounts of information through its own social media channels, the marketing
platform’s “big data” node integrates information from Haier’s corporate website and from other
sources within the company and without. The idea is to unearth cross-business insights and build
predictive models that help MEs respond to emerging customer needs. One example: alerting MEs in
the washing platform that a customer has bought a refrigerator and an oven and may be in the midst
of a remodel that will call for new laundry equipment as well.

While the marketing and manufacturing platforms do set standards—for brand visuals and factory
automation software, for example—they issue few commands. And like other units at Haier, they
have a financial stake in the success of their internal clients.

A final bit of grease on the runners of internal collaboration comes from Haier’s shared
accountability to customers. When, for example, several MEs began hearing that Haier’s smart
products didn’t talk to one another, they convened and hammered out a “grand bargain” in which
Xinchu would provide a common software platform for the company’s networked devices while
other MEs would contribute customer research and supporting technologies. The shared ecosystem,
XCook, now encompasses 100 million end users and 400 partners.

In most companies, coordination means sacrificing speed and responsiveness for greater efficiency.
Zhang believes that such trade-offs are best made by those closest to the customer, by MEs that are
free to choose when to collaborate and when to go it alone.

/
Remember Weinberger’s phrase “small pieces, loosely coupled”? The coupling of MEs is decidedly
loose but still strong enough to ensure that Haier exploits its size and scope. Turns out it really is
possible to achieve coordination without centralization.

5. From Rigid Boundaries to Open Innovation

Bureaucracies are insular. Typically, they make sharp distinctions between insiders and outsiders
and are characterized by secrecy and a reluctance to tap external partners for mission-critical tasks.
The problem with a closed system is that it doesn’t adapt—it atrophies. Recognizing this, Haier sees
itself not as a company but as a hub in a much larger network. The implications of this view are
profound.

First, every new product or service at Haier is developed in the open. For example, when the
company set out to build a new home air conditioner, it used Baidu’s social media site to ask
potential users about their needs and preferences. More than 30 million responses flooded in. Lei
Yongfeng, the project leader, then invited more than 700,000 users to go deeper and share their
thoughts about pain points and detailed product features. Unexpectedly, the top concern was the
danger of contracting Legionnaires’ disease. Minimizing that risk became a key priority and led to a
radical rethink of the fan blade.

Poorly performing leaders are vulnerable to a

hostile takeover.
Second, Haier has assembled a network of 400,000 “solvers”—institutions and technical experts
from around the world—that help the company address challenges in some 1,000 domains. More
than 200 problems are posted each year on the Haier Open Partnership Ecosystem (HOPE). Lei’s
team, for instance, asked for help in designing the blades for its new air conditioner. Within a week
the challenge had attracted several proposals. The winning design, mimicking a jet turbofan, came
from researchers at the China Aerodynamics Research and Development Center. In all, 33
institutions contributed to the development of the air conditioner. When it launched, at the end of
2013, the Tianzun Wind Tunnel was an instant hit.

/
In collaborative projects like the Tianzun, Haier creates a “pool” in which its business partners
confidentially share their patents—with the understanding that they’ll be rewarded if their
technology is used in the final product. Suppliers that contribute to the early design process also get
preferred consideration when it comes to vendor selection.

By moving its product development process online, Haier has reduced the time from concept to
market by up to 70%. Manufacturing and design nodes, user MEs, potential customers, and business
partners work in parallel throughout, starting with the earliest discussions about customer needs.
That maximizes creative problem solving and minimizes the risk of clumsy handoffs as the product
moves toward launch. While many executives view their businesses as linear value chains,
beginning with R&D and ending with sales and support, Haier sees them as value networks in which
all parties collaborate at every stage.

A third feature of Haier’s commitment to openness is its use of crowdsourcing to gather feedback on
products and defray development costs. In part this is a response to the company’s “zero fund”
policy, in which new offerings don’t get a significant budget until they’re validated by users. Take
the Air Cube, a groundbreaking combination of humidifier and air purifier. During its gestation,
more than 800,000 online “fans” offered comments on it. Once a prototype was ready, it was made
available on a popular crowdfunding site, where more than 7,500 individuals opted to buy a
preproduction model. Their feedback helped Haier further refine the Air Cube before its formal
launch.

Finally, Haier uses HOPE and other online platforms to recruit talent. Many MEs are led by people
who joined the company after making standout contributions online. Tan Lixia, Haier’s chief
financial officer, sums up the company’s mindset toward open innovation this way: “The border of
the company is not important. If you can help create value for users, it shouldn’t matter whether
you’re an employee or not.”

6. From Innovation Phobia to Entrepreneurship at Scale

There’s a reason big companies are frequently outmaneuvered by newcomers: Bureaucracies are
intrinsically conservative. As Laurence J. Peter, author of The Peter Principle, wryly put it:
“Bureaucracy defends the status quo long past the time the quo has lost its status.” To counter this

/
reverence for precedent, many companies have set up outposts in Silicon Valley and other
innovation hot spots.

Haier, by contrast, has turned its entire organization into a start-up factory. Its 50-odd incubating
MEs currently account for more than 10% of Haier’s market cap. They run the gamut from
Hairyongi, a fintech start-up that securitizes loans to small businesses—notably, Haier suppliers and
distributors—to Express Cabinets, a network of storage lockers that allows local farmers to deliver
directly to consumers in some 10,000 communities. (For more on how Haier builds new ventures,
see the sidebar “Birth of a Microenterprise.”)

There are three ways to launch a new business at

Birth of a Microenterprise Haier. In the first and most common case, an

In May 2013, Lu Kailin, along with three internal entrepreneur posts an idea online and
colleagues at Haier, set out to build a invites others to help flesh out the nascent
laptop computer for video gaming. The
business plan. (This is how Zhang Yi, who at the
upside seemed enormous. Rising incomes
and ever-cheaper technology were stoking time was an after-sales service manager working
demand for online games, while the in the field, started Express Cabinets.) Second, a
business-oriented laptops on the market platform leader can invite insiders and outsiders
were ill-suited to hard-core gaming.
to submit proposals for exploiting a white space
The team’s first step was to pore over
opportunity. Third, would-be entrepreneurs can
30,000 online reviews of gaming PCs.
Having distilled out 13 customer pain pitch their ideas at one of Haier’s monthly road
points, Lu and his colleagues wrote a note shows across China, which connect local
to Zhou Zhaolin, head of the Haier
innovators with platform leaders and members of
platform that included the laptop
business, begging for a meeting. Zhou was Haier’s investment and innovation platform.
initially skeptical. “My first instinct was to
kill the project,” he says. But then he
Every incubating ME is a separate legal entity,
realized that this really wasn’t his call. “In
making decisions,” he says, “we have to funded in part by the founding team. Haier’s
let users and entrepreneurs—not leaders, recognizing they may not be well placed
managers—speak.” He gave the team a to judge the merits of a new idea, often require a
little seed capital (1.8 million RMB, roughly
$270,000), with the understanding that start-up team to obtain outside funding from one
further funding was conditional on a of the company’s venture capital partners before
successful market test. agreeing to contribute internal resources. In a
recent period, nine out of 14 newly hatched MEs
/
 The team set out to produce a new laptop received external investment before getting
with the help of outside partners such as
money from Haier. Despite this, Haier often ends
Quanta Computer, a Taiwanese
manufacturer of PCs for Dell and HP. By up with a majority stake in the start-ups, because
December 2013, only seven months after it it typically has the option of buying out its
began, the venture was ready to introduce venture partners using a preset valuation formula.
a product. Offered on JD.com, a Chinese
e-commerce site, the first batch of 500
brightly colored and aggressively styled Like other units within Haier, incubating MEs
laptops sold out in five days. A few weeks contract with nodes for development,
later a second batch—of 3,000 units—
was snapped up within 20 minutes. Jazzed distribution, and administrative support. Arm’s-
by that success, the team members length internal agreements allow fledgling MEs to
crafted a detailed business plan and in leverage Haier’s size and bargaining power while
April 2014 received an additional 1.2
avoiding the risk of bureaucratic meddling.
million RMB from Haier, to which they
added 400,000 RMB of their own money in
exchange for a 15% stake. VC firms joined Explaining Haier’s penchant for entrepreneurship,
in subsequent funding rounds.
one VC said, “Microenterprises are like a
A little more than three years later,
reconnaissance unit—they scan the battlefield and
Thunderobot went public on China’s NEEQ
market with a valuation of 1.2 billion RMB identify the most promising opportunities. It’s
(about $180 million). With a staff of 80, like a giant search function.” Haier understands
the venture now leads e-gaming laptops in that innovation is always a numbers game. The
China and is making significant inroads
into other Asian markets. Taking a lesson only way to find that next billion-dollar
from its corporate parent, Thunderobot opportunity is to launch a slew of start-ups and
has spawned its own start-ups, which give each one the freedom to chase its dream.
include a business that streams video
games, a platform for organizing e-sports
teams and tournaments, and a foray into 7. From Employees to Owners
virtual reality technology.
In a start-up, people tend to think and act like
owners. Often they have equity in the venture,
and some will have even risked their own capital
in hopes of scoring a big win. Start-up teams also have a large degree of autonomy—and no one to
blame if things go wrong. It is this combination of upside, freedom, and accountability that gives
start-ups their edge.

/
A study of 780 U.S. companies published by the National Bureau of Economic Research explored the
connection between gain sharing, autonomy, and voluntary turnover rates, which the authors used
as a proxy for employee engagement. Turns out, neither gain sharing nor autonomy on its own had a
significant impact on turnover. But in companies that offered employees both, voluntary turnover
was less than half the rate observed when one or none of those two conditions were present.

This makes sense. If you increase employees’ authority without increasing their upside, the
additional responsibility may well be seen as a burden. Conversely, if you grant people stock without
increasing their authority, they’ll still feel like minions.

At Haier, MEs are expected to be self-managing, and their freedoms are formally enshrined in three
rights:

Strategy. The right to decide what opportunities to pursue, to set priorities, and to form both
internal and external partnerships.

People. The right to make hiring decisions, align individuals and roles, and define working
relationships.

Distribution. The right to set pay rates and distribute bonuses.

These rights come with a commensurate degree of accountability. Targets are broken down into
quarterly, monthly, and weekly goals specific to every member of an ME team. That makes it easy to
see who’s performing and who’s not. Compensation is tightly coupled with business performance.
As is true in most start-ups, base salaries are low. Opportunities for additional compensation are tied
to three performance thresholds:

Baseline. When an ME’s quarterly sales and earnings growth exceeds a base target, team members
get a bonus proportionate to the amount by which the target was exceeded.

Value-adjusted mechanism (VAM). If the ME achieves a midpoint goal between the quarterly
baseline and leading targets, the team’s bonus is doubled. At this level, employees are also
allowed to contribute their own money, typically 15,000 RMB (about $2,200) each quarter, to a

/
 special investment account. If the team hits the VAM target the subsequent quarter, that
investment produces a 100% dividend.

VAM annual target. When an ME team beats its VAM target for four consecutive quarters, it
becomes eligible for profit sharing. Twenty percent of the ME’s net profits in excess of the VAM
goal are distributed to the team, though 30% of that amount will be set aside to fund bonuses the
following year. As an ME closes in on its leading target, the profit share increases proportionately,
sometimes exceeding 40%.

This combination of bonuses, dividends, and profit sharing gives employees the opportunity for
hefty payouts. With so much at stake, it’s hardly surprising that ME team members have little
tolerance for incompetent leaders. If an ME fails to hit its baseline targets for three months in a row,
a leadership change is automatically triggered. If the ME is meeting its baseline targets but failing to
reach its VAM targets, a two-thirds vote of ME members can oust the existing leader.

New leaders are chosen competitively. Typically, three or four candidates will present their plans to
the ME team. The discussions are intense, as team members press for details on how prospective
leaders will get things back on track. Occasionally, a team rejects the entire slate of candidates and
the search process goes to round two.

Poorly performing leaders are also vulnerable to a hostile takeover. Anyone at Haier who believes
that he or she could better manage a struggling ME can make a pitch to its team. Performance data
for all MEs is transparent across the company, so it’s easy to spot takeover opportunities. If an
interloper’s plan is convincing, a leadership change ensues. This may seem extreme, but it’s simply
an analogue of the market approach to corporate control. If a company consistently underperforms,
its board will eject the CEO—or the business may be bought by a competitor who believes it can
manage the assets more effectively.

In most large companies, upside opportunities are modest, often topping out at 10% or 20% of base
pay. The implicit message from executives to employees: “We don’t think you can do much that
would make a truly significant difference to our business.” Lacking both freedom and upside,
frontline employees have little choice but to live down to these meager expectations, and in so doing

/
reinforce management’s lack of faith in their abilities. By contrast, Haier does everything possible to
turn employees into owners. It is here that one finds the deepest explanation for Haier’s track record
of industry-beating growth and innovation.

CONCLUSION
Unlike Alibaba or Tencent, Haier isn’t one of China’s new-economy superstars. Thirty years ago the
company was a struggling collective enterprise turning out products of dubious quality. Today it’s a
case study in what can be accomplished when an established company is willing to challenge
bureaucracy’s authoritarian structures and rule-choked practices. Who would have imagined that
it’s possible to run a large global business with just two layers of management between frontline
teams and the CEO?

The Haier we see today was nearly a decade in the making. The company began testing the concept
of small, entrepreneurial sales and marketing teams in 2010. A year later self-managing teams were
introduced in product units. Those early tests were instructive. At the outset internal contracting
proved problematic. Negotiations were protracted and adversarial as every unit sought to maximize
its own success. The solution? Build in a clause that links compensation to marketplace results. That
reduced friction and increased alignment, turning a zero-sum game into a joint effort to create value
for customers.

As Zhang often reminds his colleagues, it’s impossible to engineer a complex system from the top
down. It has to emerge through an iterative process of imagination, experimentation, and learning.
When asked how Haier can accelerate its transformation, he has a simple answer: Run more trials
and replicate the most successful ones faster, because revolutionary goals are best achieved through
evolutionary means.

For decades, most companies have worked diligently to optimize their operations. More recently,
they’ve raced to digitize their business models. Important as this is, Haier has done something even
more consequential: It has humanized its management model. As Zhang said in a long-ago meeting
with one of the authors of this article: “We want to encourage employees to become entrepreneurs
because people are not a means to an end but an end in themselves. Our goal is to let everyone

/
become their own CEO—to help everyone realize their potential.” Haier’s empowering, energizing
management model is the product of a relentless quest to free human beings at work from the
shackles of bureaucracy.

A version of this article appeared in the November–December 2018 issue (pp.50–59) of Harvard Business Review.

Gary Hamel is a visiting professor at London Business School and the founder of the
Management Lab. He is a co-author of Humanocracy: Creating Organizations as Amazing as the
People Inside Them (Harvard Business Review Press, forthcoming).

Michele Zanini is the managing director of the Management Lab. He is a co-author of

Humanocracy: Creating Organizations as Amazing as the People Inside Them (Harvard Business
Review Press, forthcoming).

This article is about ORGANIZATIONAL STRUCTURE

 Follow This Topic

Related Topics: Managing Organizations

Comments
Leave a Comment

Post Comment

11 COMMENTS

Andreas Engel 3 months ago

/
Bureaucracy is evil but is it really gone or is it just not visible right now since it has shifted? - Ok, some 600 VPs less,
but how to lead 3600 ME "CEOs" towards overall profitability, sustainable customer satisfaction, regulatory
compliance, brand … ? It is possible and can be done like related concepts of profit center structures show but it
comes at a price … additional coordination, arbitration or alignment (to avoid the term bureaucracy).

 Reply 00

 Join The Conversation

POSTING GUIDELINES
We hope the conversations that take place on HBR.org will be energetic, constructive, and thought-provoking. To comment, readers must sign in or
register. And to ensure the quality of the discussion, our moderating team will review all comments and may edit them for clarity, length, and relevance.
Comments that are overly promotional, mean-spirited, or off-topic may be deleted per the moderators' judgment. All postings become the property of
Harvard Business Publishing.