每天一篇经济学人 | The magic formula 神奇的公式（20...

This is the age of the data scientist. Employers of all kinds prize people with the skills to capture and analyse enormous amounts of information, to spot patterns in the data and to turn them into useful insights. But some of the most valuable figures in business need neither an analytics team nor knowledge of Python. They are simple to remember and useful for bosses in every organisation. Here is a small selection of management’s magic numbers:

这是数据科学家的时代。各种各样的雇主都很看重那些有能力捕捉和分析海量信息、发现数据模式并将其转化为有用见解的人。但是一些最有价值的业务人士既不需要分析团队也不需要Python知识。它们很容易被记住，并且对每个组织的老板都很有用。以下是一小部分管理的神奇数字:

Zero: Doing nothing can be the most valuable thing a manager can do, as the fable of Atwood’s duck demonstrates. Jeff Atwood, a computer programmer, is credited with popularising the (possibly apocryphal) story of a piece of deliberately unnecessary work that an animator did on a video game called “Battle Chess”. 

0: 什么都不做可能是一个管理者能做的最有价值的事情，正如阿特伍德的鸭子的寓言所证明的那样。电脑程序员杰夫·阿特伍德普及一个(可能是杜撰的)故事，这个故事是关于一个动画师在名为“决战西洋棋”的电子游戏中故意做了一些不必要的工作。

Aware that the higher-ups needed to feel that they were adding value by making changes, the animator gave the character of the queen a wholly extraneous pet duck. Sure enough, the reviewers asked the programmer to do only one thing: remove the bird. In theory everyone ended up happy, except the duck. In practice time had been wasted because people higher up the chain needed to justify their existence. 

让高层意识到他们在增加价值，他们就要做出改变，动画师给了女王这个角色一个完全无关的宠物鸭。果然，审稿人只要求程序员做一件事：删了这只“鸟”。理论上，除了鸭子，每个人最后都是幸福的。实际上，时间被浪费了，因为更高一级的人需要证明他们的存在。

One: This is the number of bosses people should have. In reality, matrix structures and team-based approaches mean that lots of workers report to multiple masters. According to a Gallup survey in 2019, 72% of American employees occasionally or consistently work in different teams. This approach can have benefits, but clarity is not one of them. The Gallup poll showed that those who work in a matrix are less likely to know what is expected of them, and more likely to spend their day festering in endless internal meetings. Managers in matrix structures should at least try to make their underlings feel like they have one boss, even when they do not.

1: 这是人们应该拥有的老板数量。在现实中，矩阵式的结构和基于团队的方法意味着许多工人向多个上司报告。根据盖洛普2019年的一项调查，72%的美国员工偶尔或一直在不同的团队工作。这种方法可能有好处，但不清晰。盖洛普的调查显示，在矩阵型公司工作的人不太可能知道公司对他们的期望是什么，更有可能把一天的时间花在没完没了的内部会议上。矩阵式结构的管理者至少应该努力让下属觉得他们有一个老板，即使他们没有。

Three: In a paper published in 2013 two academics tested whether there was an optimal number of claims that marketers should make for their products and services in promotional messages. They found that making three claims was best: any fewer and consumers felt they lacked enough information to make their minds up about a product; any more and they became sceptical that the claims were authentic. The “rule of three” is useful in many other settings, too, from points in presentations to pricing options for customers. One place it does not apply is in a column about magic numbers, so: 

3: 在2013年发表的一篇论文中，两位学者测试了营销人员在促销信息中对其产品和服务的广告语是否存在最佳数字。他们发现，三种广告语是最好的: 少于三种，消费者会觉得他们缺乏足够的信息来对产品做出决定；再多的话，他们就会怀疑这些广告语的真实性。“三法则” 在许多其他情况下也很有用，从演讲PPT到客户的定价选择。有一个地方不适用，那就是关于神奇数字的专栏，所以:

Ten: The number of people who should be in a meeting depends not just on what is being discussed, but also on where it is taking place. According to a survey of British workers conducted in 2021 by Nicholas Bloom of Stanford University and Paul Mizen and Shivani Taneja of the University of Nottingham, the efficiency of online meetings declined steadily as the number of participants grew. Zoom calls work best with between two and four participants, when there is less need for people to keep muting and unmuting, more chance to see people’s facial expressions and less chat-room blather. Efficiency declines until ten people or more are involved, at which point it is better to hold meetings in person.

10: 参加会议的人数不仅取决于讨论的内容，还取决于会议的地点。根据斯坦福大学的Nicholas Bloom和诺丁汉大学的Paul Mizen和Shivani Taneja在2021年对英国员工进行的一项调查，随着与会人数的增加，在线会议的效率稳步下降。Zoom通话在两到四个人的情况下效果最好，这时人们不太需要保持静音或取消静音，有更多的机会看到人们的面部表情，而且聊天室里的废话更少。在10人或更多的人参与之前，效率会下降，这时最好是面对面会议。

150: Dunbar’s number postulates that the number of stable social connections a human can have is roughly 150. First proposed by Robin Dunbar, an anthropologist at Oxford University, the figure has its critics. Some researchers reckon it is too low; introverts think it is ludicrously high. But this group size recurs in many settings, from the congregations of single-leader churches to networks of Christmas-card recipients. Companies have also found that it has significance; passing the 150-person threshold requires a change in management practices, from informal and undocumented to structured and codified.  

150: 邓巴数字假设一个人可以拥有的稳定社会关系的数量大约是150。牛津大学的人类学家罗宾·邓巴首先提出了这个数字，但也有人对它提出了批评。一些研究人员认为这个数字太低了；内向的人认为这个数字高得离谱。但这种群体规模在很多情况下都会出现，从单一领导的教堂会众到圣诞贺卡接收人的网络。企业也发现了它的重要性；超过150人的门槛需要管理实践的改变，从非正式和无文件到结构化和法典化。

There is a pattern to these numbers. In one way or another, they illustrate the risks of addition. Expand a company beyond a certain size and social bonds will weaken. Invite more people to the meeting and you will wait longer at the start as everyone dials in. Add extra reporting lines and the burden of collaboration will spiral. 

这些数字是有规律的。它们以这样或那样的方式说明了“加法”的风险。公司扩张到一定规模之后，社会联系就会减弱。邀请更多的人参加会议，你会在开始时等待更长的时间，因为每个人都要拨号。增加额外的报告渠道，合作的负担将螺旋式上升。

The idea that less is more is not new, of course. Max Ringelmann was a 19th-century French engineer who found that adding more and more people to a rope-pulling team had an adverse effect on individual productivity. The more people there were to tug on the rope, the less sense of responsibility each person felt for the outcome and the less hard they pulled. Ringelmann’s insight is still valid. Subtraction has its attractions. 

当然，“少即是多”的观点并不新鲜。19世纪的法国工程师马克斯·林格尔曼发现，在拉绳的团队中增加越来越多的人会对个人生产力产生不利影响。拉绳子的人越多，每个人对结果的责任感就越低，他们拉绳子的力度也就越小。林格尔曼的见解仍然有效。“减法”有它的吸引力。