Hello everyone! I lead the anti-spam team at Mail.ru Group, as well as several machine learning groups. The topic of this article is self-development for team leaders / leaders, but in fact, many techniques and recipes are completely independent of the role. For me, this question is very relevant, since machine learning is developing extremely rapidly, and to at least be in the topic, you need to spend a lot of time. Therefore, the question of how and what to spend time for development is quite acute.

The content of the article, of course, is not the ultimate truth, but just a description of the results of my ongoing quest, which outlines the approaches worked out for me, based on books and trainings, on trial and error. I will be glad to discuss with you in the comments.

Since I often have to conduct interviews, I ask all applicants the question: what do you expect from the leader? In addition to the standard answers (“he must set the task well”, “he must give feedback”, etc.), in 95% of cases they answer me that the leader should be an expert. That is, people expect that the leader will pump them professionally. And we, the leaders, are partly under pressure from the environment, all the time we are trying to sit on two chairs and somehow divide the time between managing people and the expert field (in my case, development and ML).

But after all, our value for the company depends on how we distribute efforts in these two areas. If you are a team lead of five people and 100% of your time are engaged in development without taking time to manage people, then the team will work unclear how. So, most likely, your value for the company is extremely small. On the other hand, if you are not doing exactly what you would like, then everyday motivation may be low. There is no feeling of happiness every day. And periodically asking myself the question, am I happy, do I do what I would like to, I can get the answer “no” and seriously demotivate.

Any good leader knows that development takes time. And now, let's say, in the evening, when everyone has already left, you are sitting at your desk and thinking what to pump around?

Example: you are the team leader of a small backend team, Kubernetes is looming on the horizon. And you have a very distant idea of what it is. In addition, the skill of designing distributed systems is below average for you. Before you are two recommended books about Kubernetes and system design. Which one to read?

Or a second example: because of your last mistake in the project, you had to rewrite the system, that is, in fact, you did a double job. At the same time, you procrastinated half the time, could not get yourself to work efficiently, so you thought less about design, about your employees, about control, etc. What to spend time on? To the recommended two-day training in Time Management or to the same book about System Design? The answers may not be obvious.

What to develop?

Let's figure out what you can generally pump. To do this, we will build a pyramid based on fundamental competencies - this is responsibility, focus on results, communication paradigm (known win-win, win-lose, ...) and dedication to work (your degree of charge for professional development). We will discuss them in more detail very soon.

Then I highlighted personal effectiveness: how well and efficiently you work during the working day, how good your time management, learning ability, planning, self-realization, etc.

Above are soft skills(to which personal effectiveness actually applies, but because of its importance, it is taken out separately). Most of these skills are related to communication. This is a fairly broad topic. As an example, I will cite people and project management typical of managers, presentation skills, decision making, negotiations, etc.

Even higher are the core professional skills ( Hard Core Skills ): databases, understanding the work of distributed systems, some basic machine learning technologies in my case.

And at the very top are “Contemporary” professional skills ( Hard Contemporary Skills ), which periodically change in your arsenal: specific programming languages, knowledge of certain libraries, databases, etc.

Development priority

An arbitrary amount of time can be invested in learning specific languages, libraries and databases. The main thing is to understand where exactly. At every moment in time, you need to clearly understand what you are doing and why. I give priority to the lower steps of the pyramid, because the upper levels depend on them, and the development of, for example, personal effectiveness will free up time for learning a language or improving presentation skills.

Long term value

Let's say you develop personal effectiveness. Imagine two situations: you either sit eight hours a day and work efficiently, or you spend four hours a day on it (we were all in one state and another). In the first case, you basically have twice as much time both for development and for work, so the lower the level, the more long-term investments you make.

If you can educate in yourself fundamental competencies, such as responsibility and focus on results, then they will be with you all your life, and no one will ever take them away. Communication skills and personal effectiveness will persist for decades, because different paradigms change periodically. I will give instant messengers and phones as an example. If 30 years ago it was possible to work calmly, no one pulled you, now messages are sent to the messenger, Facebook, etc. that distract and create new difficulties. Therefore, periodically have to be rebuilt. Active use of VR or attaching a wire to the back of your head can dramatically change our life and skill requirements.

Hard Core Skills will be useful for about ten years. For example, 15-20 years ago it was not required to be able to design distributed systems, but now it is extremely important. Some modern trends in the development will be useful for several years. Periodically, they will transfer to basic professional skills (for example, neural networks). About 5-10 years ago it was the patrimony of a few researchers, and today it is the basis of machine learning.

Based on these judgments, it is obvious to me that it is better to allocate time for mastering Time Management than System Design, but System Design, in turn, is more important than studying Kubernetes. Later I will talk about the two lower levels, since with the higher it is more and more clear.

Fundamental skills

There are four fundamental skills (as I learned from Nina Afonina’s trainings):

a responsibility;
focus on results;
communication paradigms;
passion for business.

A responsibility

Let me give you an example about anti-spam and the harsh realities of working in such a team. Spammers attack our mail service all the time. Constantly adapting to the protection system, they create such pressure that you have to constantly saw new features. As soon as you have cut something, you should immediately start sawing something new - this resembles the confrontation of a shell and armor. Such powerful and continuous pressure at the time led to the fact that we had a huge technical debt, because there was not enough time for all the necessary defense subsystems.

Let's look at the collective image of team lead in our team, which faces the above difficulties.

A person with an underdeveloped sense of responsibility (which we don’t take) is unlikely to want to rake up the problems associated with difficulties and overcoming the system. “It has always been like that, in the ass project, it cannot be fixed, we have 100 services, it always will be, I cannot change this.”

If the timlid has an average sense of responsibility, then, understanding that he can make a difference, he takes responsibility. Perhaps he will try to knock out time from his leader to do this, and he will manage to do a lot. But when such a team leader encounters serious problems (for example, a bug in the prod put mail for an hour), he hands over and tries to shift the responsibility to his colleagues: “We don’t have enough testers, we don’t allocate enough time, the admins are to blame.”

With a high level of responsibility, a person is responsible for himself and for his team. He understands that the mistake made by one of his people is also his mistake. Also, proactive behavior is certainly present - a lesson should be learned from a serious mistake and every effort should be made so that it does not happen again. As a result, the problem goes away over time.

Focus on results

And again on the example of antispam. Hundreds of millions of pictures come to our services, and for their processing we introduce neural networks, which are a highly loaded system. A team leader with a low focus on results will not be able to draw up a sound plan of action. Rather, these will be approximate deadlines and approximate goals, as a result, tasks will almost never be completed on time.

With an average level of focus on results, there is already some kind of plan and understanding of the risks. If you agreed that the task will be solved in one quarter, then most likely it will take 3-5 months. But if serious obstacles arise, the goal will change: “We incorrectly assessed the capabilities of the service, we do not have enough servers, there are no resources, so we will run the feature only for a part of users or for specific pictures.”

With a high level of focus on results, there is always a plan B. The main thing is that when serious obstacles arise, it is not the goals that change, but the ways to achieve them. At the moment, faced with this problem, we are trying to take non-trivial steps for us: ask for the purchase of servers for someone else’s budget, since ours is exhausted, take servers from someone in the current cluster and drag it into ours, and write something in return, or help in some other way.

But no matter how cool a developer or team lead you are, all your technical skills will be an order of magnitude less effective if you do not solve the problems in the project, that is, you do not take responsibility, do not complete the solution to the problem, do not do it on time and qualitatively. Any task that is not 100% completed remains with you. And this lump will grow.

How to pump responsibility and focus on results?

The gradation described above helped me a lot. With her help, I was able to determine my current position and understand what is missing. Since then, once a month, once a quarter, once a year I spend a retrospective asking myself: “What risks did I not take into account in the project? Where could I be more result oriented? Where did I try to shift responsibility to others? Was there a conflict? ” etc. I constantly try to learn something new, including pumping this skill.

Communication paradigms

It's about win-win and win-lose - how much you are focused on yourself and others. It should be noted here that the development of nature is based on the unification of some agents in order to become something big. For example, the cells in our body together can do much more than acting alone. Genes combine into DNA, animals into groups, etc. If one country goes to another war, while taking resources from it, then these resources flow from one pocket to another. And if, instead of war, trade is conducted, then mutually beneficial cooperation is obtained, as a result, everyone benefits.

If you think that win-win is not a good strategy, then I advise you to study the experience of the evolution and survival of living things. :)

Helping people in our world is not just an effective strategy - it positively affects you. Sincerely helping another person, you yourself feel satisfaction from it, because you empathize. Everyone knows how to help people, so I will focus on the non-trivial - the dissemination of ideas.

Richard Dawkins once expressed the idea that ideas are viruses. They infect people's minds and are transmitted from one person to another. Ideas evolve, undergo natural selection, mutate, etc. Successful ideas are fixed in the heads of people, not successful ones die off. “ There are no bad people, there are people with bad ideas in the head ” (c) Sam Harris.

However, good ideas will not spread themselves. So, if you learn something good, valuable, bring it to the masses. Around us there is a lot of information noise and all sorts of nonsense, therefore sound information is like a ray of the sun in the dark.

Business enthusiasm

Let's move on to dedication. In order to develop, firstly, you need to want this. Secondly, to execute it in a quality manner. To want something, it is very important that you have the resources of your body, including the brain. There are four types of resources: physical, emotional, intellectual and spiritual (goals and values). If there are difficulties with any of them, it will be quite difficult to find motivation for development. For example, without sleeping for two nights, it is hardly possible to talk about some kind of development. It is also very difficult to seek motivation to change something if you are exhausted by conflict at work. A common problem is the lack of several types of resources at once. For example, a conflict with the leader and the mismatch of goals and values with the company.

You think: “I'm going on vacation. I’ll rest for two weeks on a paradise island, I’ll be back, and everything will be fine. ” This is actually not working. Rest restores only physical resources, maximum - relieves emotional stress. But as soon as you return to work, everything is again in a circle. Therefore, resources must be restored strictly from top to bottom.

Recovery

How to do it?

Regarding spiritual resources, it is important first of all to determine your values. And be sure to write them out in priority order. A few years ago, having created such a list, I realized that my work is in the first place, although, in theory, my family, relatives and friends should be higher. Then I concluded that I did not devote enough time to them, and set a goal to communicate more with other people, with my friends and colleagues. But it’s not enough to set a goal. You need to understand how to follow it. I started planning this activity, among others, in ToDoist. For example, set the task: once a month to organize a meeting with family, friends, play poker. It is important here that goals in terms of values are a plus not only for me, but also for others (in the context of the win-win strategy described above). In this case, if you are organizing a meeting with someone, it is important that you are not the only one to enjoy this event.

To restore intellectual resources (learning new things, hobbies) you need to do what you like. For myself, I determined that development should be not only at work, but also at home. I read books, listen to podcasts. Even when I lack some kind of development due to overloading of tasks at work, I can always learn something new outside the office. Then I understand that I am not standing still. And many recognized ideas one way or another find their application both in life and in work.

Emotional resourcesFirst of all, they are filled with positive emotions. In our country, we are not very good at celebrating our victories and successes, not to mention praise. Having launched a successful product or feature, it makes sense to buy a cake, pizza, and celebrate it with the team. If you see that the other person has done something well - do not be silent, praise. Besides being nice to people, this behavior becomes contagious. Over time, your team members will begin to do the same, eventually creating a positive atmosphere.

Physical resources. Sleep, food, sports. Nothing new. But I want to discuss two extremes in load balancing.

. , , , . , . , . . - , (, ), , . 10 , , , , .
. - , , . , , , .. , (, , ), . , , . , , . , 10 , , , , .. . , , , .

Goals and motivation

Let's move on to setting goals, and also consider the existing shortcomings.

Let's say the task is to increase KPI by 10%. In our anti-spam environment this means a reduction in the number of user complaints about spam in the inbox.

In the course of moving towards this goal for six months or a year, you are always guided by the final result, and in fact, you are always in an unsatisfied state from the fact that you are still not at the final point. However, having reached the set goal, the joy does not last long, and you set new goals.

The next point: each goal must be measurable. This goal is very difficult to set accurately, since people generally do not plan well ( planning fallacy is that people think they can plan(c) E. Yudkowsky), and the future is difficult to predict. Therefore, the goal may be underestimated, easily achievable. And then a sense of accomplishment arises, then you can do nothing further and ask for higher salaries for success. Although it was possible to make efforts and do even better.

Or, conversely, the goal may be unattainable. For this reason, there is a steady demotivation, since dopamine in the brain is released as it moves toward a goal that has good chances of success. It turns out that, like, he did everything he could, and the result is bad.

There is a temptation to use the slogan “Goals for Losers” (taken from Scott Adams's book “How to Fail at Almost Everything and Still Win Big”, which I highly recommend, link) But what then is for “successful” people? For them - systems that gradually increase the chances of achieving the goal.

Let's analyze the difference using an example. Let's say you want to get more money or move up the career ladder. You can set a goal: to take the place of a boss or similar in the market. You live in the shadow of this goal, trying to do something, maybe substitute someone, or frantically looking for an analogue in the market. It turns out that you are optimizing a short-term goal that interferes with a long-term goal, instead of choosing a much more promising strategy and place of work.

In contrast, you can take for yourself a system that will make you more valuable in the labor market. Gradually developing skills and expanding knowledge, your value for both the company and the market increases. And when the opportunity to increase appears, you will be ready to receive it - such a statement of the question is not directly related to time (this is a plus), since, I recall, people are planning very poorly, including the time period for achieving the goal.

(source: Scott Adams book presentation)

The concept of the system, as opposed to goals, indicates a sense of progress (which is a stable supplier of dopamine to the brain and, as a result, feelings of satisfaction and happiness). You constantly study something and understand that you do something every day to become better in this direction. It brings emotional comfort.

The next example: you can set a goal to lose weight by 5 kg, or you can create a healthy diet that over time (not necessarily in the near future) will lead us to the result and generally improve our health.

I will give speeches as a personal example. For me, speaking is stress. I am an introvert. I like to spend time with a book or play some game. But there are motivators: a kind of altruistic philosophy - to give something useful, to spread good ideas, image, networking, etc. Therefore, instead of the goal of speaking N times in M months, I created a system for myself: at every opportunity to speak and create opportunities myself by sending applications to all suitable conferences. Then I simply execute without thinking, and suffer in the course of preparation. This can lead to one performance for the whole year, or maybe ten (as I had in 2019). None of the options will upset me, provided that I know that I have done everything necessary in the framework of the planned efforts.

Personal effectiveness

We figured out the motivation, we have adequate goals and desire. It is up to quality execution. We turn to personal effectiveness.

I perceive personal effectiveness, mainly, as the fight against procrastination.

It is important to understand that it is possible to control any state of ingenuity and productivity - for this it is necessary to identify factors that increase the likelihood of such states.

Here we again return to evolutionary psychology, to the fact that our brain is imprisoned for life in the wild, and not in the office. As Daniel Kahneman, author of Thinking, Fast and Slow ( link ), found out, people think absolutely irrationally. In 95% of cases, they make decisions instinctively, based on heuristics typical of archaic times. For example, people most often choose a president by their appearance: if his chin is more courageous, then this is a good president.

The part of the brain responsible for rational thinking is very lazy, it rarely connects and affects our decisions. If you set a goal to eat right, and colleagues bought pizza in honor of your birthday, then innate reflexes will force you to reach out to the pizza and eat a piece. It is likely that the rational part of you will not be able to influence your actions at this moment. Indeed, in order to withstand "automatic" actions, one needs to show willpower. From this we can conclude that you should organize your environment in such a way that the ancient heuristic part of your brain was very easy to make the right decisions.

What is meant here?

Are you drawn to play the console, instead of preparing for a performance? In this case, preparation should be done where there is no prefix. One of my colleagues, for example, rents a PS4 for a limited period of time (Sasha, hello again!).

Instead of work, do you want to browse through social networks? Make sure to block access to entertainment sites at the level of / etc / hosts in advance, so that it would take longer to unlock them - in these seconds you can have time to change your mind and get back to business. Try to perceive yourself as a robot, program yourself and your environment to make the right decisions.

To make such decisions (to program the environment in the right way) you need to wait for the “moment of strength”, that is, when you have slept, well-fed, healthy and capable of accomplishments (for many, this condition occurs on New Year's holidays, just like setting annual goals). For me, this moment is Sunday evening. I sit down and plan a working week. If I start to make a plan on Monday morning, when I am broken, then the plans will be the same.

The path to efficiency

It is important to understand what exactly needs to be programmed in yourself and your environment. Track which solutions work and which don’t, where errors occur. For example, on Mondays, I spend a retrospective of how I procrastinated the week before to find patterns.

At some point, I realized that the phone distracted me very much from work. As soon as you sit down and start to do something, messages and letters immediately come, everyone pulls me. So I set up the phone in such a way that it seemed to ask every time why I was holding it in my hand. For myself, I realized that I was not for the phone, but the phone for me. On the first screen, I have nothing interesting, so when I take the phone in my hand, I’ll be guaranteed not to be hooked. Messengers are hidden in a separate folder on the third screen and in no way make themselves felt. I also turned off all notifications. The screen does not turn on when I take the phone in my hand. All the messengers I check only when it is convenient for me, and not when someone wrote. You can reach me instantly only by phone call, which my colleagues and relatives know about.

The second example is informational noise . Our memory is associative. What you consume and what you have in your head greatly affects all your thoughts. At a certain point, I realized that news is mostly informational noise. There is no news on my work laptop - they are blocked at the / etc / hosts level.

In preparing this material, I specifically went to the news site (on Lenta.ru, because Medusa is blocked for me). What did I see on the first page? News of the origin of food on presidential flights. Very informative. Over time, I almost stopped reading the news. I usually learn all the important things from friends and relatives, in a sense, parasitize on them. I do not urge everyone to such a lifestyle, otherwise there will be no one to learn useful news from. But I recommend significantly limiting their reading. This will free up a ton of time. It’s better to read a book, not news and social networks.

In addition to regular retrospective, monitoring your reaction helps a lot. Improving observation, that is, attention to detail, helps today's fashionable meditation. Many, when they hear this word, represent a Buddhist sitting in a lotus position. In fact, meditation is a simple exercise for attention: you need to sit down, straighten your back, close your eyes and watch your breathing, listen to your body, to feelings. Inhale, exhale. If you get distracted and think about work, you say to yourself: “Okay, thought about work, I’m going back to breathing.” And you repeat this exercise for 5-20 minutes.

As it turned out, this exercise is infinitely difficult to perform, because a variety of thoughts climb into the head all the time. Even the ones you don’t want. Just try to sit for 60 seconds without thinking about anything. I guarantee that none of you will be able to do this, even if your life will depend on it. Meditation is a kind of gym for the brain, which trains attention and understanding of what is happening in your head, what thoughts arise, so that you can work with this at the moment of their appearance, until they capture 100% of attention.

The practice of meditation improves concentration, the ability to enter a state of flow. Now I’m less likely to sit on headphones, because extraneous sounds are usually annoying and distracting. As soon as I have thoughts that are not related to the current case, I quickly notice them and switch back.

Also, meditation perfectly pumps emotional intelligence - the ability to recognize your own and other people's emotions and work with them. Emotions are very fleeting. For example, you were cut on the road, you got angry. Anger itself will pass in a few seconds. But then you start to think: “He cut me. Now I will catch him and cut him too. ” Such thoughts start and spin the flywheel of emotions, which in turn generate other thoughts. This cycle can last quite a long time, but the difference in the quality of life when you experience negative emotions for 10 seconds or 10 minutes is very significant.

Last example: stress and excitement before and during performances. It turns out that you can learn to recognize the moment of the emergence of feelings of anxiety, until it has grown into excitement or stress, and turn your attention to something else. This practice reduced my feelings on the day of the performance to a couple of minutes.

By pumping recognition of your emotions, you simultaneously learn to recognize the emotions of other people, since the same mirror neurons are responsible for this, which is an unconditional profit for communication.

You can begin to meditate using the meditation application, which introduces the course and accompanies in this process. After meditating with the application for one year, I went to Silence Retreat - this is two weeks of intensive meditation, 18 hours a day, which had even greater effect on me. For more details and colors about the trip, you can read the links: part1 , part2 , part3 , part4 , part5 .

Manager effectiveness

The task of the manager is to maximize the result of the entire team. If you are a team leader and there are five people under your command, but you do not lead them, do not increase the results of their work, then your effectiveness is very low. I recommend reading the book “High output management” by Andrew Grove, Intel President ( link ) on this subject .

The task of the manager is to perform the most useful work at any given time, which maximizes the result of the team.

A manager's output = his output + the output of his team

If you have a problem with the technical debt, you have to rake it. If people leave the team, and you need to do management for a month, then for a while you should forget about self-development in machine learning.

It helps to understand that your task is to perform the work most useful for the job - this is your value. You stop getting nervous due to non-observance of the time balance. For example, setting up all the processes in the team so that it works like a clock, but your additional effect of supervising the team is close to zero, it will probably be wiser to transfer the team and focus on something else, where you will be more useful.

The second part of the equation is that the leader’s result is not only the results of his team, but also other teams that this leader can influence:

A manager’s output = the output of his team + the output of the teams under his influence

If you see a problem in a neighboring team and you can solve it, you should do it. For many, this idea is not to their liking, if only because people usually perceive helping others as a waste of their time.

And at some point I thought that my maximum depends on the influence on those whom I can influence, those in turn affect others, others on third ones, etc. And if I, an effective leader, develop my employees to increase their effectiveness, then I should focus on employees who prefer the win-win strategy. I can infect them with the ideas of developing other people, and they, in turn, will be able to develop other employees. That is, taking just a few hours to develop a win-win employee, I can sit in the coveted third chair and watch how the effect of my actions spreads.

As a result, I came to the conclusion that the priority in choosing between self-development in managing people or in development is shifting towards the development of employees, because spending a little time on the development of colleagues, you can get a much greater effect.

Summary

It’s worth investing in longer-term skills, because they will be with you longer and affect everything else.
It is worth building a system for self-development, and not a goal, because psychologically it is much more comfortable and allows you to optimize longer-term goals.
Programming yourself and your environment. We perceive ourselves as a robot. We understand that we are rational only at some rare moments, therefore we are engaged in programming our environment.
It is necessary to maximize the result of all those whom you influence.

I hope I have infected you with my ideas enough. You will try some of them. I think they will help you, and you will distribute them further.

PS: the post was written based on my speech at Saint TeamLead Conf 2019.

Self-development: how I did not sit on two chairs and found a third