Oh, My Code: How MAPS.ME Works

    MAPS.ME - number 1 maps for travelers. Today it is 110 million installations worldwide, the ability to add copyright tourist routes and independently change maps. How does a unique project for travelers, says its head Yevgeny Lisovsky.

    Tell me something about yourself.

    I was born in Norilsk in 1982. Moved to Moscow in 1989 He studied at the energy-physical lyceum 1502 with MEI with a bias in physics and mathematics. I have always been interested in exact sciences. At the same time he studied at the music school in piano. Then he entered the MEI at the Radio Engineering Faculty as an engineer for biotechnical and medical devices and systems, because he believed very much in the future of biotech. While studying at the institute, received a second degree - English. In the fourth year I was born the first son, it was necessary to earn money. Before that, I worked as a sysadmin for my father: I served a local network, collected computers and a little Kodil under 1C.

    I was looking for a job as a web programmer or a sysadmin. But it so happened that I found a job as an internet marketer at the international company Radmin - Remote Administrator (remote control of computers). There I began my career in 2004. Engaged in contextual advertising, SEO, PHP code. It was very cool.

    In 2010, I wildly wanted to startups. I started all bad times: I started the project “CouponBonus” (coupon service). The CEO of Buy and Bonus was then Zhenya Miropolsky, and now he is a top manager (Managing Director) of the famous international project WeWork. I was a marketing director. Then I went to another startup, the BabyBum online store for kids. This is a classic e-commerce. They raised $ 1 million at the first round. But since unit-economy did not converge well, they were unable to close the second round, and towards the end of 2011, the company collapsed. Then I went to the "liters", e-books. There is a small average check, the level of piracy is 96%. After “LitRes”, already in 2016, we met with Gursky, talked about the integration of “LitRes” into “Odnoklassniki”. This did not work out over the last couple of years. As a result, Gursky offered me to head MAPS.ME. I became very thoughtful. I had a strong motivation to stay at Lithuania. But when will there be such an opportunity - to lead an interesting international project? After all, “liters” is more about Russia, and MAPS.ME - the whole world. It motivated me a lot, and here I am.

    It turns out that you both got an engineering degree and were engaged in programming. Developing as a marketer, I didn’t want to return to engineering, to development?

    In 2004 I finished the 4th course and took up the diploma. We used C ++ to make an electrocardiogram analysis system (ECG). My class mate did the hardware, and I did the software. It was necessary to finish the product to good condition and implement certain interface features. I spent two weeks kodil and clearly corresponded to the stereotypical image of the developer: red eyes, did not sleep until 4 am, T-shirts are indelible, sweater is stretched. So much immersed in this topic that for two weeks I almost did not get up because of the computer.

    In the end, everything turned out well. Then I was not very sure of myself as an entrepreneur. It seemed to me that the world is big and scary. In fact, if we then began to develop this product, perhaps now there was a biotech company. In the US, there is the company Alivecor.com, which has implemented this idea in the form of a device for a wristwatch, which removes the ECG and gives a preliminary diagnosis. In our thesis project, software patterns were used, by which it was possible to assess the state of the cardiovascular system in just five minutes. Enough breakthrough stuff.

    How useful will be to developers, programmers to understand marketing, what is it, why is it needed?

    This is super useful. I would recommend that all people develop empathy. Technicians develop communication skills, understand marketing, its importance. And I recommend to marketers to dive into the technical implementation, at least in the form of writing a good TK. This is not a good thing for the humanities, because you need to sit down, think over all possible scenarios, write it all down. It can tire them, unlike developers who like to build algorithms, tables. And so, and others need to move towards each other. Then there will be a very good symbiosis. Such alliances give very good results from a business point of view.

    How do you implement new features? Who does the task come from? How do you measure it all?

    There is a backlog. Users write: it would be great to add this, or there is such a problem. There are bugs, technical debt, we put it in a plan and gradually fix everything. The second source, which we began to actively use in 2016, is user feedback. We conducted a fairly large survey, were surprised at how actively people answer our questions. We received a huge feedback and realized that this is a super-channel that needs to be used further. We try to conduct large-scale research a couple of times a quarter. From 300 thousand users, to whom we sent the push for the first time, we received 60 thousand fully completed questionnaires in five languages, with a breakdown into regions. But user surveys are not enough to formulate a strategy. If you ask whether users need it, they will say what they need. Get a huge range of everything

    Also in 2017, we formed the principle of DNA, which is a filter for everything that should not be done. There is a cool phrase: "it is not enough to know what to do, it is important to know what not to do." We position ourselves as the number 1 card for travelers (# 1 Travel Maps). We have more than 100 million installations, but they are very smeared around the world. We are not trying to be a local super-navigator. DNA allows us to clear a little glade and properly rank the tasks. We need to do something that will strengthen us in the map segment for travelers, will highlight the rest.

    Our positioning and detuning from competitors:

    I really liked TripAdvisor's travel stages concept: dreaming, planning, booking, experiencing. And this cycle, when your previous journey ends, you start thinking about the next one.

    We are now in the experiencing stage. People do not use MAPS.ME as an app for finding inspiration where to go. It is used as a tool to not get lost, orient on the spot. Therefore, MAPS.ME usually download a couple of weeks before traveling on the recommendation of friends.

    Gradually, I began to understand that first of all it was the food problem that had to be solved. Do not try to earn more money, but try to think, how would we be present at all stages of the tourist cycle, to do something really breakthrough. So the concept of tourist routes came to the surface. We will release it soon.

    Our high-level product strategy:

    When tourists go to travel, they can place some marks on the maps. And we, in fact, create the largest UGC-catalog of routes, which no one else in the travel industry has. There are some startups who tried to do something similar, but they didn’t take off, because it’s very expensive to attract an audience. In the travel industry a very high customer acquisition cost. And we have a source in the form of maps. Cards are the core and the heart of the service, which allows us to super-cheaply get a huge audience. We have a lot of "organic" due to the fact that people recommend MAPS.ME to each other.

    Having discussed all this, we began to prioritize the list of 180 tasks:

    • How the feature affects the retention of the audience.
    • Will the feature help to attract a new audience (acquisition)?
    • Whether the feature will bring profit (monetisation).

    This is how our task prioritization table looks like ( download template ):

    DNA is the final frontier, we recline what does not match our DNA.

    For two months we were engaged in prioritization, preparing a development strategy for 2018. There were additional factors: the audience to which the feature is now targeted, and what part of the audience the feature can reach. For example, car navigation in MAPS.ME is now actively using 10% of the audience. We know that if we make a super avtonavigator, this share will grow to a maximum of 40%. So we estimate the significance of the feature in terms of the efforts that we can spend on its development.

    Then we divided the features into thematic blocks, and inside them we again prioritized the tasks. We selected 3-4 of the most important, then scattered tasks for the year, with the distribution by quarters. Then there are iterations - sprints, lasting a little more than a month, because something unexpected always appears. Two to three weeks of development and two weeks of testing, regression and release. Approximately 10 to 12 releases per year. That's the way our planning and development works.

    This is how our business model looks like (the structure is preserved, the numbers are not real), download the template :

    You have described in great detail how your tasks are formed and where they come from. Do people at every level of submission understand in such detail how a product is created? An ordinary programmer will understand this?

    We have weekly meetings on Fridays, at 12 noon. Technical Director talks about the achievements of the technical team for the week, marketing - about business performance. I usually talk about strategy: where we go, what we do.

    So you yourself tell the whole team about the strategy?

    Yes. I think this is a mandatory part. If your team does not know where you are going, there is a feeling of misunderstanding, fear. I believe that everyone should clearly understand what we are doing. When a grocery plan and strategy is formed, it is sent to all the guys. We have no secrets from colleagues.

    How many people do you have working on the service?


    And what is the audience?

    Now we have about 105 million units. About 12 million monthly active users (MAU). It is clear that people put the application, delete. We have about 20–25 million active install base, these are people who have MAPS.ME installed, but they use it sporadically. Russian users account for about 12%, and mostly Europe - 40%: Germany, France, Spain. 15% of users are from the Americas, mainly from South. And 20% is Asia.

    How many developers do you have?

    16 people write code.

    Looking for developers now?

    Are looking for.

    What kind?

    We are now doing several web services for travel planning. Therefore, we need Python-developers for the server part.

    Do you have cartographers in your team?

    Not. Maps are created by volunteers-cartographers in the service.OpenStreetMap.org (OSM) . I call it Minecraft for Adults. Usually, everyone starts to draw maps from the houses in which they live themselves: they added a path there, a bench here. Then he went somewhere on a bicycle, recorded a track that is not on the map, and loaded it into OSM. Today in the world there are about 60 thousand. Monthly active mappers that use OSM. A sort of Wikipedia in cartography.

    How difficult is it for an ordinary person, technically not very knowledgeable, to take and add his dacha to the map?

    This is super-easy. I even loaded my eldest son. He now draws at home. OSM is an open community. You can edit the map directly on the web. There are OSM based tools where you can view your profile, progress and statistics.how you mapping compared to other users.

    There is also a mobile editor built into MAPS.ME. This is our own development, launched in June 2016. After that, the OSM community began to grow quite noticeably. The editor is very simple. It does not have the full functionality of a web editor, but allows you to edit and add information about objects, add addresses of houses.

    About a year ago a very interesting article was published in which you described a possible future, when people stop participating in the creation of maps, robots will be engaged in this. When will this turning point come and how will it affect MAPS.ME?

    Ideally, it would be good to do it in conjunction with a person. Suppose I open some area on satellite images, press a button, the neural network recognizes the contours of roads, houses. And the person then checks if this is true. On the planet there are such plots of land on which houses stand one on another, heaps of buildings, it is not very clear, this is a dirt road or asphalt road. In such cases, a person can not do. A person should help train the neural network. Recognizing satellite images alone is not enough to form an entire data set.

    For example, the restaurant is located indoors, there is contact information. From the satellite image you can not make out the schedule in the showcase. But it can be seen from an unmanned vehicle that drives down the street and removes everything around. Although not everyone hangs the perfect timetable on the street with the correct, well-recognized fonts. I think that over the next five years, cartography will be greatly simplified due to artificial intelligence and neural networks.

    Now blitz. What superhero would you like to be?



    He flies, strong.

    Mac or Windows?


    Bitcoin buy or sell?


    What book did you read last?

    "The task of the three bodies" by Liu Zixin.

    What book would you recommend to our viewers.

    “Think slowly, decide quickly”, Daniel Kahneman.

    What was the latest mobile app that surprised you?

    I really like the Peak app , these are such quiz for brain development. Various memory tasks and so on. I set my children.

    Who do you see yourself in 50 years?

    Let's start with the fact that I plan to live long enough. Over the next 20 years there will be a strong breakthrough in biotech. It will be possible, at a minimum, to slow down the aging process, and there it will get somewhere. Over the next 5 years, I see myself as an entrepreneur in the biotechnology segment. Further, it is quite difficult to plan.

    Hypothetical situation: a robot can replace you. Everything that you do is just as effective a robot can do. What will you do?

    The amendment is that the neural interface is developing. It will be possible to pump the intellect in a biological and biotechnological way. As a biological organism, you will be able to pump yourself up to a sufficiently high level of intelligence, which is fully capable of competing with artificial intelligence. The question is, will artificial intelligence have an intellectual singularity? If we talk about 50 years, about long-term plans, it is to colonize and develop other planets. I love astrophysics very much and everything connected with it. But until the issue with the engines and movements over long distances.

    Office or work from home?

    Office. Communication with people. We are very social creatures. I can periodically work from home when I need to collect my thoughts, focus. Still, the office in this case should be understood as a place where you communicate with colleagues, partners, and so on.

    The last question is provocative. What do you miss your salary?

    I have enough for everything. I want to invest more. I want to buy stocks of companies I understand. I want to invest, perhaps, not so much with money as time, into some projects that are interesting to me, like a mentor. But there, again, the question is not the salary. Now I spend the main money on children's education, 4 main items of expenditure: education, music, sports and travel. Everything else goes below priority.

    PS: all my achievements on various startups are freely available here: www.lisovskiy.ru/edu/ .

    Also popular now: