aboutarchive booksmastodontwitter

Under New Management

← Back to booksAmazonAudible

Scientific management in the 20th century has shaped how companies are managed. Focused on measurement and rationality, it has enabled much greater efficiency of industrial companies.

We inherited those management methods. But the realities of creative firms and the information age are very different from the industrial revolution. We’re only today learning what are the best ways to manage those.

Some ideas:

1. Ban email


(Note: For a book promoting supposedly radical ideas about running organizations, it sometimes doesn’t go nearly far enough. It says email is bad, but doesn’t, for example, explore the new and more efficient methods of communication…)

2. Employees first, customers second

An inverted hierarchy.

Customers can sense if employees are happy and care about their jobs. And that’s true not only in service firms where the relationship is ongoing (like consultancies). It also seems to be true in companies where the relationship is one-time (like a restaurant).

Ergo: Happy employees → good service → happy customers → profit.

Another inverted hierarchy: thinking of management as a support, not boss role. They’re there to clear the way for employees to do their work, not command their work.

3. Unlimited vacation policy

Note: I don’t believe in that one. Practice shows that an “unlimited” vacation policy generally makes people take less vacation, not more. So it’s kinda bullshit.

But consider “unpolicies” in other areas. It’s about trust and simplicity. Instead of making complex rules and processes, by default, just tell employees to use company resources responsibly. (For example, give employees a company credit card for things they need to do their job.)

4. Pay people to quit

Offer people free cash to quit instead of staying.

This weeds out people who aren’t going to be a good fit anyway.

Quitting a job involves a sunk cost fallacy. Even if it’s obviously not the right job for you, you wrongly figure that since you’ve already invested so much in the job, it would be wrong to quit now. A quitting bonus might offset this cognitive bias. It forces you to re-evaluate your decision, and if you’d quit anyway, you’ll just quit now.

A good deal for the employee (who doesn’t have to stick around for a job they hate anyway), and for the employer (who doesn’t have to pay many more months’ salary to a disengaged employee).

On the other hand, if you do decide to stick around, the decision not to take the bonus reaffirms your commitment.

So works well on both ends: weeds out disengaged employees, and engages more those who want to stay.

5. Make salaries transparent

Employers have an unfair advantage over employees due to information asymmetry over salary numbers.

By making this information public, employees tend to get paid more.

But even for employees, salary transparency doesn’t obviously seem like a good thing. It’s scary to reveal this information. In practice though, in an organization where this is the norm, it doesn’t tend to be a problem, as long as salaries are actually fair.

People really care about their pay being equitable — connected to their performance, role, and improvement. So it’s not just the absolute number, but the relative value (relative both to others’ and one’s past pay) that’s important to whether people judge it as fair.

Seeing other people’s pay be higher sucks, but not as much as when the numbers are a secret. Because if you find out when you’re not supposed to, you can’t have an honest conversation about it. But if the numbers are transparent, and the system for determining those numbers applies to everyone, you can challange the values that determine your salary.

6. Ban non-competes

Jesus, non-compete clauses are insane. No need to explain that.

But there’s evidence that banning non-competes also makes sense for the employer.

The obvious benefit is that you can attract more of the best employees (those who are priviledged enough to choose where they work).

But also, by being less paranoid about protecting your stuff, you create a denser network of ideas between companies. For example, when you study patent applications from competing companies (that don’t have non-competes), you’ll see a lot of links between each other’s patents. When someone leaves and joins another company, yes, they’ll take their ideas with them, but the same thing will happen the other way around.

In the end, cross-pollination brings more total benefit for both entities.

7. Performance reviews

Corporate annual performance reviews are bullshit — duh.

First of all, a huge waste of time to review the whole year of an employee if you want to do it right.

Employee morale and productiviy are highly correlated to how they feel about the social structure of the organization. That idea was the root of the performance review — to create a relationship between the manager and the employee. But it quickly morphed to something else.

Stack ranking: forcing reviews to follow a prescribed curve, e.g. 20% outstanding, 70% good, 10% poor performance. The theory was that by giving you an honest assesment of whether you’re meeting expectations, and offering both the carrot (outstanding — promotion), and the stick (poor — fired), you’ll be motivated to improve yourself. But in practice, this is horrible. All it does is, it makes people in the team competitive with each other, instead of collaborating. So it ends up creating less total value for the organization.

Also, people hate negative feedback about their performance even if they have the growth mindset. More precisely, people react very badly to the perception of negative feedback. (For example, being graded as “meeting expectations” when you believe your performance is “outstanding”.) The ranking is therefore not useful as a motivator.

Instead of doing annual performance reviews, have tighter feedback loops — frequest conversations with the manager. Give honest feedback, coach, show how to improve. But cut out the ranking completely.

8. Hire as a team

In most organizations, a big chunk of new hires turn out not to work.

One problem with the traditional hiring process is that you can be easily fooled by factors that only matter for hiring, but not in actual work.

It helps if the decision is not made by a hiring manager, but by the people who will actually work with the new person.

(The number of people isn’t that important. A Google study shows that by 4 people, the confidence interval is ~85%, and adding more pre-hire interactions doesn’t help much.)

Another idea is to hold “trials” (Automattic does that). So, give a potential hire a small contract for a few weeks of part-time work, and test them out in practice. That gives you a great degree of confidence someone is a good fit, and weeds out people who aren’t serious about the job. The big downside is you’re also weeding out people who could be a great fit, but their life makes a trial like this impossible.

An aside:

The performance portability of knowledge workers is limited. The individual’s performance is deeply tied to whether the whole team fits together. Therefore, be very careful about preserving the structure of successful teams.

9. Write the org chart in pencil

The basic unit of work in most modern organizations is a project, not a job. Instead of being rigid about a structure, think of ways to shuffle people around.

When you study performance of different teams, one important factor is the connectedness of teammates. And the relationship of the two follows an inverted-U curve. If people on the team are completely unfamiliar, they can’t effectively work together. But if everyone knows everyone else, the team lacks new, fresh ideas and perspectives. The most productive creative teams are somewhere in the middle: they have many people who know one another, and some people who are new.

You can apply this in different ways, depending on your needs:

10. Open offices suck


Sure, open offices do increase collaboration a little bit (like its proponents argue), but at a huge cost to morale and productivity.

An interesting hybrid: an office where you don’t necessarily have a fixed and private spot, but you have all sorts of different spaces that you can move between depending on your current needs: open-ish spaces (in smaller pods rather than big halls; easily movable desks, couches, bean bags…), smaller offices, private booths.

Note: How about no office at all?

11. Sabatticals

Every now and then, you get to really rest from work. Sabatticals are cool for an employee.

For the employer, an employee like that will most likely get back with much more energy, and fresh ideas that could only develop with some perspective. A long leave of key employees also forces an organization to develop talent on the inside and plan for succession.

Many different ways to do sabatticals or something similar, depending on the company:

12. Fire the managers

We’re not giving smart people enough credit to what they can accomplish without management if we only let them.

Autonomy is psychologically super important. If we feel like we can’t control our destiny, we’re not going to be very productive at all. If you get guidance and support, but are not actively managed, you get to choose your project based on what fits both the company and you, your skills, your interest, and the things you want to learn. You get to be excited about it, and in the end, be much more productive.

Note: autonomy (sense of control) is not the same as independence (working alone). You can have a company of interdependent people who all have high autonomy.

13. Celebrate departures

Larger organizations really benefit from creating alumni networks. On one hand, past employees can stay in touch and connect in their future ventures. On the other hand, they can bring new clients or recommend new employees to their old employer. Also, the alumni can serve the role of market research to the company.

This seems like another kind of thing where by allowing osmosis of information, you’re increasing the total benefit created for all parties.


“The old system works, so why mess with it?” A good analogy:

An internal combustion engine “works”. It has for a century and powers the world. And yet it’s only 30% efficient at converting energy from the fuel into kinetic energy of a vehicle. We can do better than that.

Today’s companies “work”. And yet research shows only 15% of workers are actually engaged by their work. Which means that for 85% of employees, their job kinda sucks, and the organization gets a worker that’s not very productive at all.

Just like with ICEs, there’s no reason to think it’s not possible to create a new culture of management that creates more benefit for everybody.

← Back to booksAmazonAudible