外部视角与反「向上管理」：如何避免企业的负规模效应（2018 码荟演讲）

The Outside View and Against "Managing Up": Avoiding Negative Economies of Scale (2018 Code Club Speech)

> By Zhang Yiming · speech at Source Code Capital's 2018 Code Club (码荟) annual meeting · Source Code Capital website (the organizer's page) · 2018-05-04 (speech given 2018-04-20) · [Original link](https://sourcecodecap.com/code-class-post/zijietiaodongzhangyiming-waibushijiaoyufanxiangshangguanliruhebimianqiyedefuguimoxiao/)

Hello everyone. I come to Source Code Capital's annual meeting every year, mainly to distill something from my own learning and growth over the past year and share it with you. Your industries differ quite a bit, so sharing about products or business wouldn't be much of a reference; I'd still rather talk about thoughts on management and talent, which have more in common. Last year I talked about how to handle the management challenges after a company gets bigger, and shared one of our company's practices, "Context, not Control." This year I want to talk about a company's negative economies of scale.

I first thought of this topic one day last year when I was chatting with Cao Yi about how companies have many economies of scale: the bigger the company, the stronger its capabilities. Then, thinking the other way round: do companies also have negative economies of scale? Actually they do. And the bigger the company, the bigger its negative economies of scale may be. Different companies have different negative economies of scale; for some companies, for example, the capacity of their environment is their negative economy of scale. And some negative economies of scale are common to all: for example, a company's most fundamental negative economy of scale is the company as an organization itself. Cao Yi just said Source Code has grown from 11 people to 50 now; as the organization grows, maybe they'll need to start studying negative economies of scale too.

On this topic, I want to discuss a few questions with you:

What are a company's negative economies of scale, and why does this phenomenon occur?

As CEO and management, how should you escape the gravity of the organization, break through the information siege and keep the ability to get real, important information?

How should a company avoid falling into negative economies of scale once it gets big, and build an effective information feedback system?

The effectiveness of the organization gets overlooked (the lost CEO and management)

The most direct consequence of a company's negative economies of scale is that the organization's effectiveness weakens, which shows itself as information failure inside the company.

Before getting into this, let me explain what I mean by organizational effectiveness. Drucker has a saying: all the results (effectiveness) of an organization happen outside the organization. An organization exists to produce external results, not to get together internally and throw a party. Yet once an organization gets big, its effectiveness is often overlooked. People no longer focus on external results and are swamped dealing with all kinds of internal matters.

The CEO's role is even more prone to this. I started paying attention to the problem through my own working state. I don't know whether the CEOs and managers here have run into the following:

First, your work is driven internally. At the start of a company, if you want to do something, you do it; that's very clear. But once the company is big, as CEO or manager, your calendar is decided by your subordinates, and you're often swamped with all kinds of internal routine work. Sometimes at the end of a day I feel, huh, I don't seem to have done anything today. Maybe I just replied to some messages or met lots of people, but actually didn't get anything done.

Second, your time is taken up by others. Early in a start-up, your time is entirely yours to arrange, but once the company is big, the time spent on management grows, and your own time becomes other people's time. Subordinates can come to you anytime, anywhere, and break your train of thought, and the things you planned to do often have to give way to "other people's things." Now with instant messaging it's even easier: messages anytime, anywhere. I don't know whether you've ever counted how many messages you get a day. If you look at each one, think about it and reply, how much time does that take over a day? The bigger the company, the more people surround you, and the bigger this drain on time.

Third, you take in information passively. Before, you might have researched the market and understood users yourself. Once the company is big, you mostly read reports, listen to briefings, look at PPTs and talk with people inside the company, getting information about the outside world indirectly. Information obtained this way may differ greatly from the real situation outside, because it isn't intuitive or first-hand understanding. Also, these understandings are very fragmented; no one does a relatively systematic input for you.

In this situation, the CEO and management also add to the company's gravity, which in turn affects the organization's effectiveness.

"The outside world may know your company better than you do"

When a company gets very big, a bad outcome may appear: information feedback fails, and the outside world knows your company better than you do.

Let me give an example. A few years ago, because of recruiting, I often met employees of a certain company, including middle managers. I found that this company probably had many problems, and later they were all borne out one by one. But I always wondered: why didn't they know themselves? Logically, being inside the company, they must have had more information channels, could look things up in all kinds of databases and could easily talk with people, while I only knew from outside. Yet the company's management seemed to understand its situation less well than I did.

From then on I kept asking myself: could the same thing happen in my own company? Later I found that it really did. In some respects, people outside the company, such as peers, rivals and new start-ups, understood the situation better than I did. And from what I've observed, this phenomenon is very common. Another example: every time investors invest, they do DD (due diligence), and after the DD they generally share the results with me. When I looked, the information investors had gathered in a month also showed, in some respects, a better understanding of the company than mine. At the time I also felt this made no sense. Most of the information the investors could get was provided by me, and the rest came from external interviews. Why would they understand the company better than I did?

In a company's early days, the CEO and management basically know about every piece of work in the company. But as the company grows, the business gets more and more complex and needs more and more roles, and you can't possibly know the situation of every position. At one point I wanted to try doing every kind of job once, so I'd know all the skills involved. But I couldn't keep it up: before I'd learned the earlier ones, more colleagues had joined, and I could never try them all. If that path isn't workable, how do you solve the problem?

I worked out a framework to analyze what causes problems like this. The conclusion: because the business becomes complex, the organization gets bigger, and three problems appear: information loses focus, "managing up," and the ego grows.

Information loses focus

As I said, a company's effectiveness lies outside the organization, but its members are inside the company. To achieve organizational effectiveness, you need to do one thing: break the external goals down to every position inside. Because people inside a company often have a view limited to their own position and role, all they can see is the work of their peers and of the people above and below them, so what they report back to you is only this work, not the external effectiveness the work achieves. The gap between the two makes information lose focus. The more someone is inside the company, the harder it is for them to look objectively at the external results of their work.

Of course, the breakdown isn't easy. For some positions, the external goal is easy to break down and the target result is relatively direct. For sales, for example, you basically don't need to evaluate; you just look at sales performance. For product roles, you may need to look at retention, DAU and so on. But evaluating what external effectiveness the work of technical roles has achieved is harder. And for finance, internal audit and similar positions, it's harder still.

"Managing up" (manage up)

The second point, which I want to focus on, is "managing up" (manage up).

In management this term, "manage up," is generally a positive one. Its original meaning is how a person inside a company finds the right position, finds out what the boss needs, works better and gets the boss's approval. But inside our company we put this term in quotation marks and include it in our cultural values as a negative term. Why?

First, let me share the motive for "manage up." As we just said, once a company gets big, information loses focus, and a lot of work can't be measured directly by external results. Then how do you evaluate work results? Mainly through the boss's evaluation, especially the boss's short-term evaluation. I've noticed a phenomenon: the less contact people inside the company have with the outside, the stronger their drive to make a good impression on the boss. Subordinates easily work around the boss's needs rather than around business goals. Why? For positions with a lot of outside contact, such as sales, the effectiveness of the work is very clear: it's sales performance. But for positions that stay inside the company, performance reviews aren't directly tied to external results; often, near review time, the boss gives the subordinate a score based on impressions, so the subordinate has a big incentive to leave a good impression on the boss in the short term. In fact almost everyone has a strong motive to manage up to influence this short-term impression.

Managers, especially the CEO, are surrounded by the organization and information, and on top of that subordinates have the motive to "manage up," so it's even harder to get real, effective information. So as CEO and management, you need to build a natural immunity to "manage up."

Let me share the "manage up" phenomena that stick in my mind; you can also think back to whether you've met anything similar. For example, briefing PPTs always have lots of pages, each listing lots of tables and figures. Why is that? Because when external measurement isn't possible, this PPT decides the impression of the work. The PPT will have a great deal of filler with no logic at all; after the mass of information has made you dizzy, you'll feel the work has been done well. If you track this PPT over several rounds, you may also find that the definitions behind its figures are quietly changing too. Today it's DAU, tomorrow MAU, the day after revenue. Even with similar definitions, sometimes it's per day and sometimes per week. In short, it picks the angle that looks best, and makes it impossible for you to track over time. I haven't seen this in just one person's PPT; it comes up very often in briefing PPTs.

Of course, investors should watch out too: when entrepreneurs present their progress to you, they may do the same.

Besides PPTs with lots of pages, there's reporting only the good news and not the bad. I don't know whether you've had this feeling in management: "we're doing pretty well," every department says so, and then you find the business isn't growing, or not growing fast enough. That makes no sense at all. Why does every department say things are good, yet the overall result isn't good? Something must have gone wrong somewhere in between.

Another typical form of "manage up" is unnecessary one-on-one communication. This is my personal feeling, and you may not agree. What's "unnecessary"? For example, for many matters or pieces of work, a text message to me of a hundred characters could explain it clearly, and there's no need to come and tell me in person. Also, I'd very much prefer they say it in the group chat, because all the relevant people are in the group, and once it's said there I don't have to pass it on to others. But I'm often asked for one-on-one, face-to-face communication.

Why do people like one-on-one, face-to-face communication? I've found two reasons. First, one-on-one can construct an information asymmetry and achieve better "manage up," because a third party could jump in and point out problems [Translator's note: that is, if it were said in a group; the transcript leaves this implicit]. When you say it in the group, with more people it's easy to get differing opinions. Actually, for most matters, unless they're especially confidential, there's no need for one-on-one communication. Second, in person or on the phone, the other side can watch your reaction and adjust their stance. For example, having said a bit and seeing you're very unhappy, they pull back a little; face to face, they can control how far they go at any moment.

There are actually many more examples of manage up. I once heard of a practice: find out which WeChat official accounts the boss reads or which Toutiao accounts the boss follows, then place PR pieces with those accounts, so the boss sees them.

I don't know how you feel, but I think that for a CEO or manager, being in a "manage up" atmosphere and facing so much optimized information is really very troubling. If the information you receive has been optimized for you, which internally we call SEO'd information, you have to find other channels to verify it, and the whole thing is very inefficient.

The ego grows

Information losing focus and manage up easily lead to the third problem: the ego of the CEO and management grows.

The CEO is the topmost node of an organization, and the biggest difference from other positions is that the CEO has no boss. There's a board, but I believe your boards don't meet often either. One problem with having no boss is that it's hard to get feedback across the board. If information has lost focus for a long time and you're being "managed up," the CEO easily falls into a state where the ego keeps growing. You often say to your subordinates: why haven't you finished this work yet? Why was this piece of work done like this? Didn't we agree on a standard? But as CEO, no matter how you do, no one talks to you like that. Especially when the company's business is getting better and better, your problems are even less likely to be exposed. But with your subordinates, even if they do very good work, you still remind them not to be complacent. For example, when Douyin was growing well for a while, I was thinking whether I should remind them, and specially check whether there was any arrogance.

Of course, it's not only the CEO; management is also prone to a growing ego. As I said, as CEO or management, your channels for getting real, effective feedback easily fail: as soon as what you say makes a bit of sense, someone will give you a like; as soon as your decision has a bit of logic, someone will agree. Gradually you stop listening to differing views, ignore excellent practice outside, and when you face negative feedback you may even get irritable.

A growing ego also shows itself in some concrete ways. For example, inertia in using people: preferring people you know, being unwilling to get to know talent widely, and being unable to evaluate and compare excellent people inside and outside the company objectively. But in fact, just as with a recommendation engine, you need open recall to find the best creators; a small circle and a fixed perspective can never find the best choice.

To sum up: information failure inside the company, subordinates optimizing for their bosses, and the CEO and management feeling good about themselves make up a company's negative economies of scale; these are the disadvantages of big companies relative to small ones. There's a common saying, "that's how it is once a company gets big"; I think that's an attitude of inaction. Saying it means you don't intend to overcome the gravity of the organization.

If we sum up the problems above, they really amount to "self-hype."

Examples of "self-hype" causing problems for companies are common. For example, in 2007, when Jobs launched the first iPhone, Nokia's CEO at the time seemed dismissive. He said, "Apple won't have any effect on Nokia, because Nokia has focused on making phones for many years and has product lines meeting every price point and need, while Apple has only one product." Reportedly he had never used an iPhone. In 2007 Nokia held first place in global phone market share, but three years later it couldn't hold on.

I believe every company has "self-hype" to some degree. How do you crack this problem? I've thought of three points, corresponding one by one to the problems above.

The outside view

The first point is to use an outside view to solve the problem of internal information failure.

What is an outside view? I understand the outside view as treating the subject as an object. For an individual, it's treating yourself as someone else; for an organization, it's treating your own company as another company. For example, forget that you're this company's CEO or management, go back to being a user and experience the product; then you can actually get the most first-hand impressions.

The second path is user feedback. For example, as everyone knows, Amazon's Bezos pays great attention to users' email feedback, and their corporate culture has "customer obsession," putting the customer first. In essence this is also an outside view. Companies today basically all do user feedback; usually it's compiled into a user feedback report, but I don't think that's enough. Why? Because such reports tend to be summaries of problems; they're very abstract.

I once read a book that said if you really want user feedback, it should be direct, preferably face to face; video works too. For example, when one of our product features was hard to use, people inside the company would say, hmm, it's not great, let's change it when we get the chance. But when we played everyone a video of a user who tried several times, couldn't get it to work, got angry and swore, everyone felt it had to be fixed right away, because that's very vivid feedback.

Other outside views include peers, the media, competitors and so on. Competitors are a very good channel for external feedback. Only your competitors are likely to keep attacking you, keep finding your loopholes and keep exposing your problems, which gives you a continuous, powerfully motivated outside view. Attacks from outside peers are a good hedge against internal "manage up."

I read a book about Huawei in which Ren Zhengfei said they want to keep competitors around and not let them have too bad a time. I think that makes a lot of sense. Without competitors, generally a company's arrogance and problems keep growing; keeping effective competitors is good for the organization's health.

Former employees are also a very good outside view. While employees are inside the company they have to "manage up," but once they leave they don't, and former employees often have plenty of complaints. If you can bypass the inside of the company directly and get input from the outside view, I think that's a very effective process.

Corporate culture & tools

The second point is to use corporate culture and tools to reduce "manage up."

What I just described, direct contact with the outside, is one way to solve information failure. Internally, if information can be passed on more effectively, outside information can also be fed back better. So we've made "dare to say what you really think face to face, dare to admit mistakes, don't pretend, don't worry about face" a requirement of our corporate culture, and internally we're explicitly against "managing up": expose problems and seek truth from facts. If this kind of corporate culture can be achieved, outside information can flow more directly inside and won't be blocked.

Internal tools are also a path. For example, I said earlier that one form of "manage up" is that the data in briefing PPTs often quietly change definitions to show the work's results from a better angle. For this kind of problem, one thing we do is build a good BI (business intelligence) system and install all the SDKs (software development kits) at the front line of the product, bypassing intermediate departments, so that no one can modify the data and the CEO can see the results directly. Generally, data that subordinates have hand-picked for their bosses tend not to be very reliable. This sounds like a small problem, but it's actually very common. Using a BI system to let information punch straight through, with no human link in between, is one solution. Of course, this only works for a small part of internal data; much other internal operating activity can't be structured or can't be punched straight through.

Besides BI, internal communication tools are also very important. A good internal communication tool lets anyone in the organization, inside the organization and in a suitable form, feed outside information back in. For example, we have an internal "Bytedance Circle" where colleagues can give feedback and vent directly, for the CEO, business leads and relevant parties to see. Increasing internal information flow and internal transparency also reduces the distortion of information and the chances of "manage up." Those are some of our solutions at the level of corporate culture and tools.

Shrinking the ego

The third point, about the CEO and management, is shrinking the ego. On this topic, the best way of putting it I've heard is what Wang Xing said before: shrink your outlook (your "three views": of the world, of life and of values) down to the size of an atom. An atom is small and hard. When your ego shrinks to the size of an atom, first, it won't block your view, and second, it won't easily be pierced. Conversely, if your ego is very big and inflated, it blocks your view: you see your ego first, and many facts you can't see. Maybe a rival writes an article attacking you and you can't stand it.

Shrinking the ego is the foundation for the CEO and management to take in information fully; otherwise it's easy to shut out negative feedback. Especially when the company gets bigger and the business gets better, the organization has a strong platform effect, and the team and the CEO easily amplify the leverage of the organization, treat external causes as internal ones, and let that affect their egos.

To sum up the ways to crack a company's negative economies of scale: attention to the outside view is the outermost layer; the company's internal culture and tools are the second layer; and the innermost layer is the ego of the CEO and management themselves. This is also the most central layer, and it often causes some key problems.

Of course, besides the methods above, there's a simpler method: don't make the company so big. But I believe Source Code Capital invested in all of you hoping you'd become unicorns or super-unicorns, so that method doesn't seem very workable.

That's my sharing for today. Thank you all.

Copyright: © LookMinds. Our own edited text and translation, not a substitute for the original; rights in the original belong to the author and the platform. Please see the original link.