Where to find and how to choose timlid

Prehistory

Hello! My name is Vitaly Sharovatov, I have been working in IT for 16 years. Now I lead the frontend in Badoo. It includes two teams that develop and support the desktop version of badoo.com , the mobile version of m.badoo.com, and many other projects. Yes, the desktop and mobile versions are made by individual teams. :)

Two and a half years ago, I came to Badoo as a developer, eventually grew to the team leader, and then, when it was decided to transport the Desktop Web team to London, I became the head of the direction.
')
Last fall, at Codemotion Milan, I gave a report on growth from a developer to a timlid (and wrote an article about it on Habr) and about what unexpected moments I had to face, and now I’ll tell you how I went from the lead to the head of the direction. selection and "cultivation" of tmlid in one of the teams (Mobile Web).

Why do we need leads and where do we get them from?

I needed tmlids for very simple reasons - to ensure the manageability of a team of 25 people and to maintain the efficiency of employees while reducing the number of “entry points” to departments.

There are two classic options for obtaining a new timlid: hire or assign. But we choose the third - “grow.” First, it’s part of the Badoo culture - always helping your people grow; secondly, each developer has a good history of his work (I will tell below about why this is important).

Why are we not following the path of destination? Well, probably, many have heard, or even have seen, how they choose in the Timlids just the most intelligent engineer and how this can turn out. :)

A light version of this scenario happened in Badoo some time before I joined the company: one very competent engineer was offered to guide a small team, and for some time he managed it quite well, but it became harder and harder every day and demotivation grew and grew up. The problem was simple: the person did not want to offend anyone or seem too harsh. Therefore, in situations in which the application of the directive approach was required, he did not succeed at all in rigidly demanding something from the employees. In the worst case, such a mistake with a choice can lead to a serious demotivation of the engineer himself (and it’s good if he succeeds in returning to engineering work!) - after all, in the opinion of the person, he was entrusted with a new position, but he failed. Fortunately, in that situation everything ended well: the developer was very loyal to the company and honestly told the manager in time that he was not working, and the manager, in turn, listened and removed this “unbearable burden” from him. Could be much worse.

I have always believed that it is better to approach the selection of the team leader and the subsequent steps on the other hand, minimizing the risks and consistently taking a series of actions, which I designated as “Selection — Assignment — Support and Control”.

Selection: how?

Conscious choice - we minimize the risk of personal sensations influence on the correctness of choice, we really choose the most suitable candidates. All people are different, and the choice of team leaders should be based on certain principles, not personal preferences.

We minimize the risk of failure - we do not negatively affect the motivation of people (both candidates and other team members).

And, of course, we estimate the time costs that will be needed in the case of each candidate. If a company needs a tmlid in a month, we determine which candidate to choose: one that is almost suitable, but remains average, or one that you need to spend a lot of time (perhaps more than a month), but who will eventually become a good manager?

Selection: when?

As soon as you become a team leader, you need to start to pick up a replacement for yourself.

In my experience, this is a very good principle - as soon as I become a team leader, begin to plan the development of my subordinates, thus creating for myself a temporary “safety cushion” for the future.

But with the replacement-replacement story is slightly different. In general, it seems to me, one of the primary tasks of any manager is the search for a deputy; the work of the department should not stop if you suddenly became ill, decided to take a vacation or quit.

But for me, an unexpectedly pleasant bonus turned out to be that, when there is a ready “deputy”, you can temporarily entrust him with the management of the department, and to immerse yourself in some important project.

And, of course, with the "cultivation" of the deputy-future timlid, another person appears with whom you can beat or discuss different thoughts.

Selection: cards

What are cards for me? A simple document of an almost free (but uniform for all) form, which describes all the key things for a manager about an employee. Naturally, no one should be given access to them.

Advantages of card management:

1. A deeper analysis of the current situation and its clearer formalization.
2. Reducing the likelihood that some moments / stories / patterns will sink into oblivion.
3. Simplify the analysis of human progress and monitor the speed of its development.
4. The “objectification” of the process is a clearer separation of one’s personal relationship to a person and his professional qualities, the history of his work and development.

This is an abbreviated version of the card, much had to be omitted, but I think the structure is clear: qualities with marks, motivation, roles, goals.

I will briefly describe the structure of the card:

• current leve l - the current level of employee development in the company. Badoo has a level system similar to Google; Your company may have a completely different “ladder” (for example, the factories in the factories are taken);
• plan - my vision of an employee development plan;
• qualities - arbitrarily evaluated (my scale is from one to five) skills and qualities of an employee (for clarity, I consider it extremely important to indicate working events confirming the assessment);
• motivation - my ideas about employee motivation. Every time I review or update the cards, I try to re-evaluate the motivation. In my opinion, this is generally one of the most important points of the card, because in order to build plans for the development of an employee, it is simply necessary to evaluate his motivation;
• roles - current or planned employee roles in a team;
• goals - goals set for the employee (or me), based on his motivation, problems / successes and business goals of the team.

About the scores and scoring I will tell below.

Selection: scoring

Inside the cards, I do some scoring of candidates and update the results with each new task. I don’t know if it’s correct to call it scoring, but I’m used to using this term, meaning by it a constant reassessment of qualities important to the team leader.

• Business orientation - how pragmatic a person is when the goal is a result for a business, and not, say, a level of satisfaction with the beauty and elegance of the code. Timlid must first of all be guided by the principles of control, not personal comfort. A simple example: if a business needs to validate a concept within an A / B test, there is probably no point in covering the test group code with unit tests, even if it is customary to cover everything with tests, and the developer likes to write tests.
• Communications. In the Badoo culture, communication has built a lot, and the ability to communicate effectively with everyone - server users, testers, tops, designers - is simply necessary.
• Level of motivation In my opinion, a tmlid should have a strong internal motivation, because turning from a developer into a tmlid is not a transition from a step to a step of a career ladder, but jumping onto another ladder, which stands alone, is a control ladder. Without strong motivation, any development - and development is always a way out of the comfort zone - will be very difficult.
• The history of successful / unsuccessful projects. This also includes the reasons for the failures of the projects (in order to later check the following projects for the presence of the same failure patterns).
• The level of informal leadership. The leader must be able to lead employees, to motivate them to achieve goals. In addition, the informal leader of the team will be easier to take as timlida.

Selection: sale

Choosing a team leader, it is important to understand who will be interested in management. I call this selection stage a sale.

During regular meetings (and we have tmlids spend them with every programmer every two weeks), it is worth assessing a person’s interest in another career as a leader: show what you are doing yourself and why, and see if an employee lights up.

For example, you can ask questions from the series “Sylvester, but do you see that Mario has problems with terms? Why do you think so? How to help a person? ”And see how a person reacts, is it interesting for him to try to understand the problem.

There are two types of answers: “What are you asking me managerial questions ?! I came here to write the code. Can I go on working? ”, And“ Oh, damn, how interesting! Probably, Mario couldn’t claim time from server servers normally. Maybe we need to come up with some other planner format to solve this problem? ”It is immediately obvious that in the second case, the person’s eyes are burning, he is ready to work with people, and this work is interesting to him.

An additional bonus is the greater transparency of the reasons for which you make certain decisions, which leads to an increase in the level of trust of subordinates. After all, if they see that such questions are being asked, it means that the manager is thinking about them, the interaction is established.

Selection: projects

The project work consists in the fact that I assign large and very large tasks to the candidates for tmlids, in which many workers are involved. The way the candidate handles such a load perfectly demonstrates the level of his managerial qualities: the very business-oriented, result-oriented, level of communication skills, ability to plan and demand fulfillment of plans.

It is very important that when working on a project, other developers informally "obey" a candidate for a timlid, because he is not even higher than them in the management hierarchy, he is not able to formally demand something and direct tools of motivation. However, even if in such a situation, developers are productive in working with a potential lead, there is a high probability that under conditions of formal subordination, productivity will not decrease. But if already at this stage one can see the negative impact of interpersonal conflicts on the process, perhaps this is a signal that the candidate is not yet ready for leadership.

What else is convenient in the project work - it is easy to choose the project necessary for one stage or another: both from the point of view of the team (for example, you can put a hard person to the empath and see how he will work with him) and from the point of view of technical complexity ( maybe, throw a project-idea, where it is not even clear which way to “dig”, and the result is needed in a week).

Another advantage of the design approach is the low risk of failure. Even if the project fails, it will not greatly affect the motivation of the candidate. To motivate other employees who participated even in a failed project, this is also not very bad: “Well, Seryoga blunted, failed the project,” and not “Damn, and then work with him? Well, no, better now I’ll quit. ”

Selection: Deputy

As I said before, you’ll have to look for a deputy. So this task needs to be solved in any case, and it is also useful from the point of view of checking candidates for the Timlide.

1. To leave a deputy in his place, you need to transfer your routine functions to him, and for this, they need to be structured and documented.
2. It is not necessary to wait for your vacation - you can delegate your work for a while to the deputy (and you can start preparing the report yourself and at the same time monitor the work of the deputy :)).

In addition to the standard metrics-qualities from the card, a lot becomes clear about the person right away:

1. How easily he adapts to a completely unfamiliar type of load (and flexibility and speed of adaptation are some of the key skills of a manager, I think), how his personal work is organized, whether he manages to do everything, if he doesn’t have time, tells the manager about problems (there are whether transparency).
2. How thoroughly he performs the routine work: does he have enough corrosiveness and attentiveness for this, or will he have to invest resources in coaching in this area too. The work of a manager always implies routine actions as well. But for some people they are given very difficult, you have to literally force yourself. If a person can not cope with routine actions even for a short period of time, perhaps this is also a negative factor that should be taken into account.
3. The level of informal leadership in a team: whether the way a person works as a deputy differs from how he worked on project work. It can be seen as perceived by other employees.

After selection

What to do after successful selection tools?

At this stage, as a rule, there remain some number of candidates, and by comparing their scoring, it is possible to single out one (it is necessary to assess the risks, the amount of time costs). In fact, we are ready to assign tmlida now.

At regular meetings with team members, it is imperative to ask how they worked with this person as a project leader or deputy, prepare them for an appointment, and eliminate potential expectations and offenses.

Purpose

Having worked with the expectations at the previous stage, I am ready to appoint a team leader.
At the regular weekly meeting with the team, I officially appoint him, tell him why I chose this man, I declare officially that all questions are now being addressed first of all to them.

I also work carefully with expectations: I’m saying that I’m not going anywhere, and I’ll be monitoring everything that happens, at first helping in solving all the issues.

Support: development

After the appointment of the newly minted timlid help will be needed.

I think it is right to apply classic situational management at all stages, implying low professional qualifications (after all, a person has never been engaged in leadership work on a permanent basis before this point) and high motivation.

1. First, the team leader will have to “lead by the hand,” using a practically directive approach, say to him: “Do as I do” and explain why.
2. Then you can go to coaching, asking: "How would you do and why?" And advise on all issues.
3. After that, it is already possible to release the Timlid to float freely and go to the liberal approach “I need these results; please do. ”

Of course, you always need to keep your finger on the pulse and regularly help Timlide.

Support: control

Incorrect actions of a team leader can cost companies much more than incorrect actions of a developer. Therefore, it is necessary to work out all the moments much more thoroughly, eliminating even the slightest misunderstanding in communication with the team leader (if the developers have such moments “aligned” to each other, then the team leader is alone over the team).

And you need to precisely control the strategy, and not every byte of his work.

What to do?

1. At meetings with the team leader, when discussing any issues, constantly check his understanding of what was said: “Tell me how you understand”, “Explain in your own words,” etc.
2. Constantly “uncover the toolkit”: “Look, I brought you to the solution of the problem with leading questions, but you yourself solved it. You see how my coaching works for you - it will also work on the guys. ” It would be nice to add a personal example.
3. Allocate more time to meet with Lid, than previously spent on communication with each developer. It takes me at least one and a half hours a week to communicate with the tmlids, plus a bunch of small sync discussions.

Support: help

Implying that a team leader is highly motivated, you need to keep in mind the high probability that a person will try to be as efficient as possible and begin to process.

Burnout control is needed: do not require less, but ask to do an analysis of the time spent. I did it this way: during the day, I recorded 15-20 minute slots, “tagged” them, and then sat down once a week and analyzed what my time actually was spent on.

It is important to discuss with team leaders the prioritization of tasks, to determine what is urgent, what is important, and what is urgent and important ( Eisenhower matrix ). It is necessary to convey to the person that there is nothing wrong with the fact that he does not know something or does not have time - you need to teach the team leader to analyze the consumption of limited resources (time) and rationally approach their spending.

The team leader needs to set more strategic goals than the developers, thereby accustoming him to a longer planning horizon.

The cost of unreasonable waste of time by the team in case of a wrong decision of the team leader is quite high. For this reason, we constantly remind the team leader that our credo is pragmatism, and any solution should answer the questions “Why?”, “How much?” And “What are the risks?”

And one more important point: you need to bring to the timlid that gentleness leads to the fact that people sit on the neck, but the sequence in the development of employees - to the pleasure of results and growth.

. , «» - . : , . .

, , :

1. : , .
2. — .
3. , , — , , .
4. , - . , .

Thanks for attention! , — .