为何不赞成控制人力成本（2016 码荟演讲）

Why I Don't Favor Controlling Labor Costs (2016 Code Club Speech)

> By Zhang Yiming · speech at Source Code Capital's 2016 Code Club (码荟) annual meeting · Source Code Capital website (the organizer's page) · 2016-04-23 (speech given 2016-04-22) · [Original link](https://sourcecodecap.com/code-class-post/zhangyimingweihebuzanchengkongzhirenlichengben/)

The business advice start-ups need differs by stage and by industry, so I want to share something universal and relatively long-term. After thinking it over, the only thing is talent, because the role talent plays in a company doesn't change.

Recently an article swept through everyone's WeChat Moments, a story about wolves and rabbits. When I first read it I thought it made a lot of sense: everyone wants their team to be a pack of wolves, working very hard. But think about it carefully and the questions appear. Where did the rabbits come from? Is this whole intake of people rabbits? Can wolves and rabbits even turn into each other? If there are many rabbits and few wolves, can you replace them? Can you tell the wolves from the rabbits? So it's not about simply venting or voicing an expectation [Translator's note: the transcript has 发现 here, apparently a slip for 简单 or similar], but about thinking through the questions behind it. That's why a company's talent mechanism is the most crucial thing.

On company growth and the talent challenge, Netflix's famous "Culture & Values" deck has an analysis, which I'll introduce together with my own views. In an early company, the business shouldn't be complicated. Because at first a company only does product and technology, not marketing, PR, media partnerships and so on, the business is very simple. But as the company grows, the business gets more and more complex, so you have to hire lots of people, and the talent pool gets diluted. Then chaos starts to appear. What do you do? Many people propose: let's set processes, write rules, create systems. That's the conventional solution. Once you emphasize process, getting anything done in the company slows down; many big companies are like this, and a company will be eliminated for lack of innovative spirit. What do you do?

So I'll answer this with a chart. The vertical axis is business complexity, representing the diversity of your company's business and how much its businesses need to coordinate; the red line is talent density, the density of excellent people; the black line at the bottom left is process and rules.

There are four possible outcomes:

1. The company doesn't become complex and keeps a small, lean team. But that's no use. A platform company or a large company certainly has to absorb as many factors of production as possible and become a very powerful system. Only a system with high throughput is a good system and can create great value. So keeping complexity low and the team small is not the solution for a company that wants to do something big.

2. A fairly common one: as the company's complexity rises, it raises the rules too and adds processes, using more process to prevent mess and problems. This solves the immediate problem but actually brings a big loss. From the point of view of the department setting the rules, to have fewer problems it will certainly want processes and rules as detailed as possible, but this weakens many possibilities of finding the optimal solution, because the situations we face may call for elasticity and flexibility. If there are many restrictions, employees won't want to look for the optimal solution. Especially when the industry undergoes a major change and the company can't move forward on inertia, the accumulated rules, processes and systems are an especially big problem.

3. There's also a rather bad case: without even process, things become a mess. Companies with process usually aren't described as messy, only as slow and rigid.

4. Another way is to raise talent density: add people with a sense of the big picture, very good values, and well-rounded knowledge and ability [Translator's note: the transcript has 只是 here, apparently a slip for 知识, "knowledge"]. Basically, I think the two dimensions at the bottom balance each other. If the people you hire understand things extremely poorly, the company's systems have to be very detailed. But facing a group of high-quality people, you can make the rules very simple, simple enough to be a few principles. People coordinate by following principles rather than hand-holding rules, and can even act just by knowing the "goal" combined with common sense.

Of these three dimensions, the easiest to increase is usually rules, because there are always companies whose rules you can refer to. Generally, when the industry is relatively stable and the model doesn't change, adding rules is OK, because rules fix the way colleagues coordinate. But in a dynamically changing industry, rules create problems: what they constrain is flexibility. So the reason many companies later run into trouble is that after the business grows, they set rules and the company moves forward on inertia; but if the industry suddenly changes and big internal adjustments are needed, the rules become shackles.

We think an industry like Toutiao's is an innovative one, especially facing continuous challenges and changes in these few years, so we think we should reduce rules and keep the organization flexible to fit the development of the business.

So we believe that as a company grows and its business increases, the core key is to make the density of excellent people outpace the increase in business complexity. Our company sums this up as "do challenging things with excellent people." Challenging things are really those whose complexity keeps rising, and they need to be done with excellent people.

How do you make talent density outpace business complexity? A talent mechanism has three main points. First is reward, including short-term and long-term reward; second is growth, that people can grow in the company; third, that their inner life in the company is happy and they find the work interesting.

How do you pick the best people? What makes you able to hire such good people? Today I'll mainly talk about the first point: apart from interest and growth, I think the core is still an effective incentive strategy.

First, offer the best ROI. We often see the term "labor cost." Many companies treat talent as a cost to be consumed. Thrifty CEOs especially think: I found this person very cheaply, great. But actually, compare with America: labor costs in America are especially high, and Chinese talent who go to America all see their pay double or triple. China, India and Cambodia have low labor costs, yet America still develops best. The core reason is that America gets better returns by deploying excellent people, so the core key is not to look at cost but at return and output.

The core of a company is to build good allocation, allocate good factors of production, give the company the highest ROI and give each person a good ROI, so a company's core competition is in its level of ROI, not its level of cost. As long as ROI is good, the more you pay, the better the return, just like investing. So we've always told the HR department that we want to pay top of the market. We actively require HR to benchmark market pay at least once a year and keep our pay leading the industry. Of course, if labor costs are very high, that in turn requires the company to deploy these people well and make the most of them, but that is exactly an enterprising posture.

Second, reward should keep a high enough ceiling to attract, at any time, top talent who create super value to join and strive. Over the past year or two in interviews, candidates have often said that Toutiao has been developing for quite a while and the best time to join has passed. We find that very frustrating (though I used that persuasion tactic myself when poaching people early on). If good people won't join bigger companies and choose start-ups instead, the later competitiveness is limited. Early companies generally use option systems and give engineers a high percentage of options, but later you certainly can't give out a high relative percentage, because there are more people. How do you solve this?

I think options aren't the most crucial thing; options are just the possibility of excess returns, of financial freedom. So the core is really whether you provide excess returns, whether you give people the chance at a step-change in reward. So we think the focus should be on raising the share of year-end bonuses, and we've told people inside the company that we hope truly outstanding people have the chance to get a year-end bonus of 100 months' pay. Then we let them know that whenever they join Toutiao, the reward can be very, very high, and the platform resources are very good, which is more competitive than going to a start-up.

I've noticed a problem: up-front stock options mainly reward not employees' business ability but their investment ability. A person's reward basically depends on when they joined which company and whether they took more cash or more stock. Someone very good at the job may not be well off financially and ask for more cash when joining; later, even if they do the job very well, their reward will be much less. Where conditions allow, I actually strongly encourage putting more of the incentive after the fact, at year-end, and turning more of the incentive into something tied to personal contribution rather than investment judgment.

Third, set pay fairly and rationally according to job level and performance review. We did statistics and found that, compared with performance, pay often carries various kinds of premiums: an acquaintance premium, a newcomer premium, a seniority premium and so on. So inside Toutiao, monthly pay is set by job level, which represents a person's stable output in their professional field. We don't let business managers set pay; they only set the job level, and we don't let them refer to past pay either. Whatever someone's pay was at their last company, it shouldn't influence and distort the job-level assessment. HR then makes an offer based on the job level combined with supply and demand and the intensity of competition at the current stage.

At the annual review, we actually want it treated as a re-interview: if this person were joining again, what offer would you make them, and at what pay? If you would make someone an especially senior offer, you should consider raising their pay substantially, because they've grown fast. This should be judged against start-ups in the industry: if they could take on the role of a start-up CEO in the industry, their package should be completely different.

Likewise, if someone is still rather lacking, would you still make them an offer, demote them or let them go? Because relationships inside a company aren't only business relationships; between people, more acquaintance-premium factors arise, and we want managers to look at this very rationally. We even want managers to imagine: if a subordinate told you they were resigning for a better job, would you accept it fairly easily, or would you be very sorry?

A CEO should be an excellent HR person. If you treat the company as a product, it has three main inputs. The first is money, capital input. The second is opportunity input, information input: what's happening in the industry, what needs to change; this tests the CEO's judgment. The third is talent input. The company's output is profit, services, products, and between input and output it depends on the quality of the inputs and how they're allocated. Whether capital is used effectively and talent is used effectively: that part is management. Inputs and management determine output, and once the company's business direction is set, the most important input to consider is talent.

Finally, besides the CEO doing HR, HR also has to do HR well. The current state of the industry is that the bar for HR is low. I think HR isn't only recruiting; it's also taking part in managing the organization, helping the CEO and business managers recruit and allocate talent. It really needs to think deeply about the company and its organizational capability. I'd even say human resources really means the understanding of human resources. If you accurately understand your business goals, derive from them an understanding of your roles, and understand the talent in the industry very well, then in fact the industry's talent is all yours, because talent flows. If a company can't understand talent correctly, its talent isn't really its own. It often happens that someone doesn't perform very well at one company, then goes to another company or starts a business and has great success; that shows they weren't your company's talent, because you didn't understand and use them correctly. If I set a very high standard for HR, I think it would be being able to write a book like How Google Works. If HR has no understanding of how to organize and how to mobilize people to produce efficiency, and only does routine work like recruiting, it's actually still very far from being an excellent HR.

These are some of my understandings. Our company is also learning and working hard, and I hope to exchange ideas and progress together with you all. Thank you.

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.