Almost fired. How I built Yandex analytics department

    My name is Alexey Dolotov, I haven’t written to Habr for 10 years. In part, the fact is that when I was 22 years old, I started building the Yandex analytics department, then he headed it for seven years, and now I came up with and build the Yandex.Talents service. The profession of analyst provides a lot of opportunities. The main thing is to start right - for example, at the School of Managers there is now a set in the direction of analytics.

    I decided to tell how my career was developing, and give some advice to those who want to “ignite” in this profession. I hope my peculiar experience will be useful to someone.

    The only semester of the university and the beginning of a career

    By the time I entered the university I was good at programming, I even wrote my shareware product (a word from the past). It was a cataloger of laser discs. Winchesters were still small, not everything fit on them, so people often used CDs and DVDs. The cataloguer read the file system of the disk, indexed it and collected meta-information from the files, wrote all this to the database and allowed it to be searched. On the first day, 50,000 Chinese downloaded the product; on the second day, a crack appeared on Altavist. And I thought I made a cool defense.

    I entered the ITMO University of St. Petersburg, but after one semester I decided that I already know how to program, that I’m learning more quickly in the process of work, and therefore I went to Norway to freelance. Having returned, I organized a web studio for two with a partner. He was more responsible for business and documents, I - for everything else, including the technical part. At different times, up to 10 people worked for us.

    Yandex in those years held the so-called client seminars, one of which I allegedly penetrated as a journalist. There were performances including Andrei Sebrant, Zhenya Lomize, Lena Kolmanovskaya. After listening to how they say, I was impressed by their non-standard thinking. The best way to approach someone in terms of professionalism is to start working with him. Therefore, at that moment - I was 19 or 20 years old - I re-invented my whole life, decided to quit the not very successful web studio and move from St. Petersburg to Moscow to get to Yandex.

    I could not do this immediately after the move. The department, which I was trying hard to get into for some reason, knew that I had tricked into the seminar and even tried to get a certificate of the Yandex.Direct course. By the way, this certificate could not be issued to me for a long time. No one expected that anyone other than the main audience of the seminar would take the course. This story seemed strange to my future colleagues, and Yandex did not take me then.

    But Mail.Ru quickly took five interviews in two days. It was by the way - I already ran out of money after the move. I was responsible for all search services, including GoGo and But after a year and a half, I nevertheless moved to Yandex as the manager of sorcerers (extradition items that answer the user's question right on the page with search results). It was the end of 2008, about 400 people worked at Mail.Ru, about 1,500 at Yandex.


    I must admit, at first, Yandex did not work. Four months later, I was offered to look for other options in the company. In fact, they fired. I had some time to search, but if I hadn’t found anything, I would have to leave. Until then, I did not work in a really large company with a complex project management structure. Not guided, not enough experience.

    I stayed, got a job as an analyst of communication services: Fotok,, but most importantly - Mail. And here a combination of managerial skills (walking around people, agreeing), grocery (understanding where the benefits are, what users want) and technical (applying programming experience, independently processing data) was very useful to me.

    We were the first in the company to build cohorts - to study the dependence of the outflow of users on the month in which they registered. Firstly, we fairly accurately predicted the volume of the audience using the resulting model. Secondly, and more importantly, it was possible to predict from it how different changes would affect the main indicators of the service. Yandex didn’t do this before.

    Once Andrei Sebrant came to me and said - you are doing well, now we need the same thing on a Yandex-wide scale. "Do the department." I replied: "Good."

    The Department

    Andrei helped me a lot, including sometimes saying - you're an adult guy, sort it out. There are no typos, this is also help. I really needed more independence, and I began to do everything myself. When a question arose to the leadership, I tried to think first: what would the leader answer me with this question? This approach helped to develop faster. Sometimes, due to great responsibility, it was just scary. A turning point has come: I, from a person solving a problem, have become responsible for the development of a large piece of processes. The number of services and the services themselves grew, they needed analytics. I was actively involved in two things: hiring and mentoring.

    They often came to me with questions, the answers to which I did not know. Thus, I learned to solve almost any problem with some accuracy, based on a very limited amount of data. It's like in What? Where? When? ”, Only there you have to give the right answer, but then there may not be the right answer at all, but it’s enough to understand which way to dig. I began to struggle with many cognitive distortions (perhaps the most popular among researchers and analysts is confirmation bias, the tendency to confirm my point of view), I developed “invariant”, “quantum” thinking. It works like this: you hear the condition of the problem and immediately imagine all possible and impossible solutions, automatically “resolving” these branches and understanding what minimum hypotheses need to be checked in order to “solve” as many of the most probable branches as possible.

    I also taught the guys what I did not know myself. I learned the first basics of statistics from interviews that I conducted. Then he began to teach leadership, although he had just become a leader. There seems to be no greater incentive to sort things out properly than to explain it to someone else.


    I began to help analysts grow: I told everyone that I would work with him, and he should work independently with the service team. At the same time, I asked uncomfortable questions. An analyst comes to me and talks about the tasks that he is currently doing. Further dialogue:

    - Why are you doing such tasks?
    - Because they asked me.
    - What are the most important tasks for the team now?
    - I do not know.
    - Let's do not what they asked for, but what the service requires.

    The following dialogue:

    - They do this.
    “What are they not doing?” What did they not take into account, what did they forget to think about?

    I taught the guys not to take tasks until they understand what the customer really “hurts”. It is important, together with the customer, to “rehearse” the scenario of how the analytics result will be used. It often turned out that the customer did not need what he initially asked. Understanding this is the responsibility of the analyst.

    Such is the philosophy of “good partisanism” or “partisan product management”. Yes, you're just an analyst. But you have the opportunity to influence the course of the entire service - for example, through the correct formulation of metrics. Formulating metrics and goals on them is perhaps the main tool of the analyst’s influence. A clear and transparent goal, decomposed into metrics, each of which is clear how to improve, is the best way to direct the team to the right course and help it not to stray from this course. I promoted the idea that all my guys should interact cross-service and thereby form “hydrogen bonds” inside Yandex, which tended to stick out at other seams.

    Search share

    In 2011, we investigated the reasons for the change in the Yandex search share - it was difficult to prove the influence of each specific factor, and there were many of them. One Friday, I showed Arkady Volozh a schedule that I had not been able to draw up for a long time and finally made up. Then I came up with a “method of freezing factors,” which allowed me to highlight the influence of browsers with a predefined alternative search. It was well read on it that the proportion is changing precisely because of them. Such a conclusion at that time did not seem obvious: people have not used such browsers very often yet. And yet it turned out that the predefined search greatly affects the situation.

    In those days, the phase of my active communication with Volozh began: I began to devote more time to the search share. The concept itself appeared - analytics of a share or “phacapack” (sharp changes in a share were often caused by someone’s fakap). It was then that Seryozha Linev appeared in the team, in the future one of the key Yandex analysts. Together with Lyosha Tikhonov, another excellent analyst and author of Autopoet, we helped Seryozha grow and create an invaluable expertise around the detection and analysis of complex anomalies. Now, if there is any incident affecting the share, the administrator on duty will immediately know about it with all the details. It is no longer necessary, as then, to convene a dozen analysts and spend several days investigating the causes. We can say that now in this regard we have the era of spaceships, and then we dragged carts.

    Arkady was always very interested in the share. He began to often call and write to me when anomalies occurred on the devices of the search share - even if I was not related to the causes of these anomalies. Perhaps he continued to call me, as it helped. And I just knew who to call next.

    By the way, there is a newsletter on Yandex for non-working questions, and when I asked someone to lend me a badminton racket, it was Arkady who was the first to respond.


    Probably, it is here that it is appropriate to tell how I - albeit not for long - worked with Ilya Segalovich. According to the chronology, it was necessary to tell about this before: oddly enough, I worked with him while still at Mail.Ru.

    The fact is that the search at that moment worked on the Yandex engine (only GoGo, another Mail.Ru project, had its own engine). Therefore, as a manager of search services, they gave me the contacts of several Yandexoids. For technical questions, I called either Tola Orlov or Ilya Segalovich. To my shame, I had no idea who these people were then. Out of hours, Ilya’s phone number was easier to call, but on the contrary, in the afternoon. I was wondering why it is so rare in the workplace, I thought - what kind of developer is this? But when he answered, he very politely and informatively helped me in the shortest possible time. Therefore, first of all, I called him.

    Later I found out who Ilya was, even played badminton with him as part of a large group of colleagues. When settling in Yandex, I tried to remember what I had time to tell him. Ilya, indeed, by all external signs, was an ordinary good person without any kind of star disease.

    There was a case, we ran into Ilya in the elevator. Ilya, in a wild excitement, arranges for me an elevator pitch, showing the screen of her phone: “The future lies with this!” During the time in the elevator it is impossible to figure out what exactly he means. But you notice how much a person burns, and you don’t understand whether this is crazy or genius. Probably both.

    There are people whose ideas live in me and make me better. Ilya is one of them.


    Many of the current analytics departments in Yandex are now headed by my guys. There were several reasons why I decided to switch to something else after seven years of managing the department.

    First, Yandex turned into a group of companies, and the need for centralized analytics disappeared. Secondly, with such a large department, too much administrative work has become. And thirdly, I wanted to make decisions and take full responsibility for them. One day I wanted to come home and tell my wife: "This is what I did."

    Therefore, I created the Yandex.Talents service. We try to reinvent the search for work and hiring employees. We are now taking only the first steps, but I see a huge potential in us. The classic idea of ​​a job board in an era where machine learning and drones ride around the streets everywhere seems outdated. It's time to start using smart algorithms to help both job seekers and employers.

    Before, I always explained to people in services how to do their work, believing that these arguments are based on analytics and my expert opinion. But work on Yandex.Talents has shown that I am often wrong. The truth is born between people - a simple statement, which, however, must be felt. In addition, the creation of a startup required to immerse yourself in the business, and now I believe that the product analyst should first of all study the financial model of his product. If you don’t understand what your key business indicators are made up of, how can you help your team achieve them?

    What a cool analyst needs

    The analyst must be able to do a lot, but there are two main skills that allow you to truly “ignite”.

    First of all, a phenomenal ability to deal with cognitive distortions is required. I advise you to read the article "List of cognitive distortion" on Wikipedia, an interesting and useful reading. All the time you catch yourself thinking how much this list is about us.

    And secondly, no authority should be recognized. Analytics - it's about arguing. You first prove to yourself that you yourself are wrong in your conclusions, and then you learn to prove that someone else is wrong. On one of the days of August 2011, the Yandex portal for some time worked intermittently. It was Friday, and the next Monday was huralthat I led. Arkady came, cursed for a long time. Then I took the word: "Arkady, now I will proceed to the hural, perhaps." He says - no, there will be no khural, let everyone go to work. I replied that I would not let the company work all week in such a mood. He immediately agreed. And we held the hural.

    These qualities will be useful in other areas, especially if you are a leader.

    Also popular now: