"When art connects to craft": publishers of online media about technology, AI and life
We talked to the publishers of large Russian media about new technologies and formats in the media, AI, blockchain, demand for quality, fact-making and much more.

The article is based on the decoding of the round table from our November conference " Content 2018 ". Speakers - Maxim Khryashchev ( Pikabu.ru ), Sergey Baryshnikov ( Bigpikcha.ru ), Evgeny Volnov ( Cyber.Sports.ru ) and Roman Hudonogov ( Setka.io ) shared their experiences and gave a number of valuable tips to budding authors and journalists. Does it make sense to open your own media - read under the cut.
- Several years ago, media based on the blockchain appeared in which the monetization process was implemented: people get tokens for posting, likes, etc. How relevant is this? Is there such a future for such media?
Maxim Khryashchev: For content to be of high quality, the user must have intrinsic motivation. It seems to me that for money it is difficult to write a good post. Another point that I noticed, after watching a bit for Voice (Golos.io): the content is influenced by the cost of cryptocurrency. When it grows, users are happy, they have many posts on this topic. The course falls - all upset, which is reflected in the content. I think this is not very interesting to a wide audience.
Sergey Baryshnikov:I do not see in the market successful implementation examples with tokens, although many platforms are experimenting with this. Once upon a time, LJ introduced its tokens, now Yandex.Dzen is trying to pay the authors money. Wins the one who first set up this work and will be able to offer a really tangible income to its authors. But when this happens is unknown.
Yevgeny Volnov: A super-perspective undertaking, but only as a research project. If you read the "Voice", then you will immediately understand everything. People discuss the whole blockchain mince, how to scale it ... The process is like in some kind of aquarium: you can see, laugh somewhere, and be glad that we are wrong.
Roman Khudonogov:Of these blockchain pieces, I liked Brave browser the most, although this is a slightly different approach. There, the user can pay for content that he likes. When you install the browser, you get 10-15 BAT on your account - this is like the currency of Thailand, only crypto. The more often they visit sites, the more they earn. If a user more often (compared to other resources) visits, for example, on sports.ru, then most of his cryptobates go there, to Zhenya. This model is more interesting, it motivates authors and publishers to make really interesting content. And a good scheme for the distribution of financial resources.
- Let's talk about systems that advise additional content to articles. There is such a service as relap.io, which allows you to make personalized recommendations for users, taking into account their interests and some statistics. Another example is Echobox, which positions itself as an intelligent online posting. Question: how advanced are recommender services in general, and how advanced are recommender systems on your sites?
Maxim Khryashchev:We tried one recommendation system, I will not call it. As a result, she began to send users the most CTR-th records. But CTR does not guarantee quality, and we received a lot of negative feedback. Often there were very funny headlines, which I am even ashamed to voice. We have abandoned this system, but we use our own. We have the content shown in the tape, and the user can interact with him - like, vote for him, comment, etc. So it is easier and more understandable to him than with a third-party recommendation system.
When using such services, you need to understand why you are doing this, what goals you are pursuing. And assess whether it really helps somehow: it increases the viewing depth, affects your income or other indicators.
Sergey Baryshnikov:If the media has the resources to develop its own analytics services and recommendation systems, then no external system can compare with them. I respect the guys from Relap, we tried several times to work with them - but still, our internal recommendation system showed the best results, and we stayed with it.
Yevgeny Volnov:You need to use these systems to make money and put your own. As we did and it looks like everyone else. I also respect the guys from Relap, and all the guys who do something with their hands. But the recommendation systems remained, it seems to me, somewhere in 2013-2014 and did not move forward. You can do so much in this field, and no one does. Although these are obvious things. If there is someone who is engaged in recommendatory systems, come to me: I will tell you how we all can make money.
Roman Khudonogov:I also respect everyone, this is important. We had a different experience: first we made our own system, then we tried someone else’s. In the end, we realized that ours was still better. I am talking about Look At Media projects. In general, the idea itself seems a bit flawed - to recommend to readers simply abstract posts that he is most likely to click. The media has many different tasks, including information, education, and entertainment. But recommender services now perform only the last task, ignoring the educational function — in my opinion, extremely important for the media. If you read about some phenomenon - political, sporting, then it would be great to go over the recommendations to similar material. So that you, for example, dive into the topic, understand or discover something new in this direction.echo chamber , when you are in some kind of vacuum. And modern recommender systems are dragging us more and more there. And diversity is awesome.
- How do you think, will artificial intelligence come soon into recommendation systems?
Roman Khudonogov: There is a cool post on Facebook: if you see ML on a slide, then it is written in Python, if you see AI - in PowerPoint. That is, AI - this is too abstract, I guess. And the logistic regression written in Excel is quite ML and quite AI. So the recommendations already have ML, AI and everything else. It depends on how you prefer to call it.
Yevgeny Volnov:Recommender systems have two tasks: to narrow the person’s horizons or expand it. It seems that all this time we have been going to narrow it down. It seems to me that in the 2019-20-ies these recommendatory things will start to leave the agenda. People need to show some other things, and not just what they are likely to click on. Agree, it is strange to recommend a person alcohol, if he already drinks alcohol. Perhaps it is better to show him that there are many other pleasures in life?
Roman Khudonogov: Most likely, so soon they will not leave. The example of Chinese newsfeed Toutiao, which has 72 minutes per person per day with a 200-million active audience, confirms that business is moving in this direction. So the alcoholic will still be shown alcohol.
Maxim Khryashchev:Zhenya proposes to broaden the mind of readers. But the “or / and” recommendation system is very difficult to do: it does not yet understand how to show something new in order for the user to like it.

From left to right: Anton Polyakov, moderator of the round table, Habr's studio content manager, Pikabu publisher Maxim Khryashchev, Bigpikcha publisher Sergei Baryshnikov, publisher of sports.ru Yevgeny Volnov, and co-founder of Setka service (Look At Media) Roman Hudonogov.
- What formats of presenting materials are interesting and relevant today? Does something new, promising, for which people cling?
Maxim Khryashchev: I did not actively study this question, but it seems to me that the main thing is in the essence of the content, and not in its packaging. In my opinion, the format in this sense is secondary.
Sergey Baryshnikov: I do not see anything new. The problem with the media right now is that nothing revolutionary is happening. A certain fatigue of media advertisers is connected with this. Therefore, they periodically go to the brand-media, to experiment with bloggers and so on.
Yevgeny Volnov:You can recall the Yandex narrative, although it is a weak example: a good product, but it’s difficult to call it a trend ... I think we don’t know how to entertain people, speak some complicated topics in simple language - but life becomes harder and harder. Moreover, we have a very suitable country for this. I like when the media can talk about complex things in the format of quizzes, but at the same time do not slip into some kind of tin. We are on sports.ru, I am constantly talking about such a project as ranker.com: look, drop dead model, very simple. I do not remember other examples when the content would be repackaged into some kind of interactive, which a person can perceive in half a minute.
Roman Khudonogov:I found the answer while driving here in a taxi: a study by Reuters and the University of Oxford, which is called Time to Step away from the 'bright, shiny things'? The guys interviewed 39 leading journalists from 17 countries on the topic “what to do next”. It turned out that for the last five years, there have been throwings from trend to trend. Type: “Oh, storiz! Let's make a story, it's very cool! .. ”“ Quartz launched on Facebook Messenger ”- you sit there, suffer, read the news in the messenger, you don't understand why - but it's cool! Conclusion - it would be good to return such formats as journalistic investigation, narrative, and concern for the reader. This is cool, in my opinion. In general, when there is a trend, a wave of advanced technology, then at first everything rises, then - the bottom. After all suffer, and then everything starts to grow again. Here we are moving along such a J curve funnel.
Now it’s more interesting for me not so much the format as the specific publisher. The fact is that recently there are interesting alternatives to the classic media. By classic, I mean online media. Books and magazines are already somewhere far away. Now brands began to create more and more incredibly cool media. These publications are outside the model “let's accumulate an audience and sell its attention to advertisers”. They perfectly reveal topics, especially professional ones. Now I don’t know a single traditional media in which I could read many cool articles on a topic that interests me. But it seems to me that over the next 5-10 years we will see a fundamentally new media landscape, where the brand will stand on a par with the publisher. He will also become a publisher and will often solve the reader’s problem better. The fact, that brands are manically focused on User eXperience, they cannot just say: “Buy, please, from me! I'm cool! ”They need to tell the story so that the person also associates this cool message with the brand.
For example, the Intercom brand is such a thing that appears on the right of every third site and is terribly annoying - publishes a cool magazine Inside Intercom. They write not about how cool they are, but how they make the product, how they organize the support service. This is a super cool trend! The second example: I do not know any media that talks about extreme sports cooler than Red Bull. And if you ask the audience what kind of media is the best in Russia, it tells about money, how to save money - most likely, the majority will call the publication of a brand that we all know. And this is really cool. It is possible that some brand will soon release a political publication. It seems incredible, but I think it will.
- What new distribution channels have appeared lately? What did you experiment with? What channels have you gone well or caused interest?
Maxim Khryashchev: Experimented with messengers, tried Weiber and Telegram. The same groups appeared in Vaibers as in the Telegraph. They differ from old ones in that they are displayed among the list of your personal contacts. But still we did not go. It turns out well with Telegram: they scored 164 thousand subscribers, advertising is sold in the group. After some time, we may pay back investments in the development of the channel. Very well monetized channel in VKontakte. But Facebook does not pay off, but we have historically not developed a relationship with him.
Sergey Baryshnikov:In 2017/18, many media outlets hooked on the Yandex.DZEN traffic needle. Probably, everyone will be hurt in 2019 to jump off of it, as it was four years ago with Rambler. Yandex.Dzen, in my opinion, is now “number one” by the way it pours traffic. And then further - who has learned to work with him, he can monetize it.
Yevgeny Volnov:The word "distribution" has crammed. I hope in 2019 the frequency of its use will fall. So far, nothing new has appeared - and thank God: we would have to hire people, explore, stuff cones. And we are already tired of this. Everything is changing rapidly, someone constantly sets experiments on us. They are somewhere successful, somewhere not. What to do, you have to live with it. Although everything has turned into some kind of bytovuha, which is just boring to do. I agree with Roma about the media: in the media it becomes crowded that someone comes to you - advertisers, platforms and so on - and all the time tells how to do it. And brands do not have these restrictions. There are other frames, but they allow you to do something that classic media simply cannot afford because of some patterns. Media turned into brick factories: Here is such a distribution channel, let's ... The question of distribution is part of this factory for the production of an absolutely standard, uninteresting product. This is terrible, we must move away from it.
Roman Khudonogov:We began to forget that your website and your application is incredibly cool. You have direct access to your entire audience at once. And on Facebook you accumulate a group of 150 thousand, publish a post and have access to 2% of your audience. Distribution needs to be done, otherwise you will lose. But own content hubs in the form of their applications and sites - that's what's really cool! You can collect data, recommend and understand. No one can do it better than you. And the most important thing is that you are within arm's length of this very audience. For example, you can send an email to contact with the user. You do not need to rely on the mechanism of distribution of attention, which uses a particular service. I wanted to post a message, and they say to you there: “You can forget it in your group for money.
Yevgeny Volnov: There are about a hundred of our own mobile applications in different languages in our company. I know how much we spent on them and how much we earned, so I want to tell you: do not make mobile apps, especially if your media is not sports.
Roman Khudonogov: He specifically says so, I assure you: “Guys, please do not go to any platforms! Do not do the media, it will end badly ... ”

- How much did the reader's attention spread through these channels?
Maxim Khryashchev:We had this story with one of the development engineers. When he met his girlfriend and talked about his work, she asked: “Are you working in a group on VKontakte?” That is, she didn’t know about our site in principle. So due to additional sources of traffic, we acquire a new audience, which does not know our original sites, sites, applications.
- With the massive advent of bloggers, the quality of content has sunk. Is there a demand for a better product? And how good can user-generated content (UGC) be?
Maxim Khryashchev: When we started, the quality of the content was the most important factor for me. I judged by myself: why I will use the resource? And I decided that I need to go every day and see quality content. Somewhere blog content and sags in quality, but still there are some very cool materials. I like the same VKontakte experimenting with the Longrid. In general, it is very difficult now to offer content to the user, and the main struggle is underway for quality. Therefore, it seems to me that, in absolute figures, the volume of high-quality content, on the contrary, increased.
Users can make good content, but you need to understand their motivation: it definitely will not do something worth 10 rubles - or how much you can earn there in Golos. With the right motivation, the quality of UGC will be. Especially if by quality we mean not the design, but the essence of the content.
Sergey Baryshnikov: Content quality is a subjective value. I will argue about the fact that bloggers have reduced the quality, on the contrary: take the same Dudya, who raised the bar on YouTube for everyone and for the media as well.
- Yes, there are good examples. But generally, on YouTube, the quality of blogging content is very controversial.
Roman Khudonogov:The quality has not deteriorated, just more channels, communications, people. Take the millionaire channels: there were, for example, 100 bloggers - it was 12,000. Probably, some integral quality has decreased, but in general, everything remained at the level.
Yevgeny Volnov:I don’t like this content option like quality. Let's draw a parallel with art: we do not argue in the gallery about whether a quality or low-quality picture. It can be boring, uninteresting, with lots of borrowing and so on. But I can not say anything about its quality. This is in any case some kind of statement. Even a single letter is a statement that may be interesting in context. And if the bad texture, but the text of 20 thousand characters - who will read this? It can be cool and high-quality analytical content that no one needs. Because the author does not understand what people need, what is now important, how to convey your emotions. I am for any content, therefore I have no complaints about UGC, any comments, etc. The main thing is that everything should be done with talent. Then it's cool.
Roman Khudonogov: I believe in quality content - in terms of the existence of such a term.
Again, returning to brands and to a request for quality content: we want to give our readers maximum benefit. I really like it when the guys do the work for the reader, when they try to figure out a topic. So that you do not need to look for any other sources, and when this post is really cool decorated, when you are pleased to consider and read it. I would call it quality content, and it, in my subjective opinion, can defeat the mentioned cloths of the text, not very interestingly written, and so on. If we go back to some retro media, then there is just a book on an interesting topic - and a cool book written about the same. The second one wins a lot, so high-quality content is awesome. I personally have a request, I want high-quality content.
Maxim Khryashchev:I really liked Zhenya's analogy with art. Indeed, by what parameters do you like the picture, how would you rate it? Probably, this is the author’s style, the plot itself, the size - it can be all over the wall or small ... Similarly, you can evaluate the content: how it is designed, how beautiful the photos are, whether it is easy to read, whether there is humor, how facts are picked up, and so on. That is, there are a number of parameters, and it is already possible to judge them collectively. It turns out that there is a certain scoring system inside of us, according to which we evaluate how much we like or dislike this content.
- We know that people can be wrong. Sometimes posts that have a fatal error collect a million views and hundreds of likes. And no one will pay attention, if somewhere in the 15th comment someone writes: “Guys, things are not really so ...” Do we need any kind of billing system? And in general, how UGC can be reliable?
Maxim Khryashchev:Yes, there is such a problem. Some users post unreliable content, others refute it, write refutation or require, for example, to delete a post. On the other hand, in a publication without a UGC, there can be an author who will write inaccurate content - and no one will give him a feedback that this content is unreliable. We work with this on the basis of feedback of users. If they show us, the support service, some facts that refute the post, then we delve into it, study it. According to the results of the check, we either remove the disputed content from the site into a specific section that no one sees, or we edit it, or do something else with it. Users have a serious demand for fact-tracking. This is a very important indicator of content for you to read, and this is true.
Sergey Baryshnikov:As a consumer and, on the other hand, as a person who is close enough to money in a company, I consider factual billing to be an absolutely harmful thing. This is some kind of lobbying, manipulation, etc. First, it can only look like an ecosystem solution: different players sat down and agreed on some general rules like: “We need to improve the ecology of our information environment, let's invest a couple of tens of millions of rubles and our brains to solve this a problem for society. " Can you imagine such a situation in Russia? Me not. Secondly, factchkking is a good way of manipulation, including for some regulatory bodies. That is, to any media that has stumbled somewhere, they can come and say: “Guys, we’ve spoiled you, you’ve become some bad, now we will punish you ...” We see similar examples. Who speaks about factbooking at all, how about some super-big problem? Facebook and similar players are practically monopolists. Not sure they want to bring honesty into our lives. In addition, our system is still capable of self-regulation. The same Navalny is an excellent fact-tracking system: fun, interesting. I think these things are enough. Plus it is technically very difficult to do this.
Roman Khudonogov: In my opinion, this is a question like personal hygiene. I wash my hands, but to go out and shout about it is not very cool. I agree with the guys that a good solution is factbooking, based on some kind of feedback loop: users complained, you checked.
Yevgeny Volnov: Yes, there may be some kind of hybrid stuff here. At the same time, we, the media with large reach, need to think about how to teach people some kind of consumer culture, to explain many things to them. And factual billing should be created inside. There is definitely no one right way; it can happen differently for everyone.
- When articles for journalists will begin to write artificial intelligence machine? How realistic is this?
Maxim Khryashchev: I think these articles will not be very interesting. It is necessary to invest some kind of individuality in the content so that other users with a different personality would be interested in reading it. So I do not really believe in it.
Sergey Baryshnikov: On the contrary, in a year or two we will see simple materials, news from some kind of automatic media. Probably, they will arise in our field. But how deep they step further - no one knows.
Yevgeny Volnov:Some kind of hybrid models will definitely appear. Artificial intelligence can help a person put the right word, determine which of the two synonyms is more clickable, or choose the most clickable from several headers, send a post to Facebook or Yandex.DZen.
Roman Khudonogov:During the last Olympics, the robot company Byte Dance, which makes newsfeed Toutiao, composed and wrote news on sports topics. They were well structured, understandable, and not easy to say. To create each text, it was necessary to look at several sources, to understand who beat whom, to make a small podochka. There were a lot of these kind of publications, and it really worked. That is, in formats where the structure is clearly visible, robots are already writing. That's right, let the journalists do other things - more creative.
- What advice can be given to editors and authors? What technologies and services - from new - would be useful for them? And in general, what is the main thing in the content?
-Evgeniy Volnov: Do not pay attention to anyone! Do not listen to anyone. Formats? Hammer! Examples, including Dudya, with whom I worked in the same company, sat in one meeting, saying that one should not drive oneself into the framework. Yura never tried to do something like “this is to make money ... to say this and this ... and to make it look expensive.” It would be trite and uninteresting. In our time, when so many opportunities - both technological and in terms of topics - you just have to go and do.
Roman Khudonogov:I am presenting Setka service here, so recommending it to someone is not very cool. But still I want to say that design is a very important thing. If you have created a certain story, article or something else, do not forget: the correct packaging in the form of a cool structuring of the text and everything else - the work done for the reader. An important thing that is often missed, because of this, we organized our small enterprise: working in the Look At Media publishing house, we came to the conclusion that you can tell stories more interesting, better. We are trying to change this industry a little bit.
Yevgeny Volnov: Maybe there is no money. But then there is an opportunity to make a real revolution. Spit on the money, go there and do super cool things! Now there is weak competition, there are few who do very cool things. This field is free.

- At the recent RIW, the media unanimously criticized Yandex.DZen for not giving them enough traffic. Today you mentioned that Zen is paying authors money. How exactly? Now, there seems to be only tapes of the coolest Russian media.
Sergey Baryshnikov: Each author, upon reaching a certain number of subsidiaries - seven thousand or ten - can insert Yandex.Direct blocks and receive some money. Major authors receive, in my opinion, several tens of thousands of rubles a month.
- Recently watched Habr's admin panel. On average, a post with 50 thousand views has 100-150 reactions (plus or minus karma). What are the new ways to involve users in feedback - so that they press something, express their opinions? For example, Medium has claps, you can pat ...
Roman Khudonogov: It seems to me that the reaction is an atavism. If you can understand the behavior of the reader, like it or not, why ask to press it on claps? Therefore, most likely, they have become smaller.
Yevgeny Volnov:When we force a person to explicitly respond to the content, he becomes an accomplice. This is not the way to know what he thinks - as if it were important to us. But this is a way to involve him a little bit in content creation. I like the widget from Disqus, the comment service: you can read the article and display your mood as you know - you know how in Sberbank, how you were served. It may seem strange, but it's fun. Plus, all sorts of things like “You read it - and what do you think?”, “Who is right, this one or the other?” I tell about the fact that we entertain poorly. Only in this direction there is still a lot to be done, and no one does anything. By the way, I do not understand why.
Roman Khudonogov:In UGC projects, this resonates: it is cool to involve a person and give him not even an illusion, but the ability to somehow influence. This is cool, I agree.
Maxim Khryashchev: The guys said about the user who leaves a reaction. But it is equally important to think about the author of the content, who wants to get some comments, reactions, advantages. It is important to assess the audience - well written article or bad. Regarding statistics, it is important to clarify how many of the 50 thousand who viewed were registered users, how many logged in. After all, not the entire audience can leave this reaction.
Yevgeny Volnov:By the way, we talk so much about all sorts of donations to the authors. Does anyone know a tool that could be easily embedded on the site and that would allow without any hemorrhoids to directly pay the author, so that it would be some kind of scalable system? I dont know.
- What information resources do you personally read? Western examples are especially interesting.
Roman Khudonogov: I really like the MIT Technology review, especially their electronic journal. They began to publish the publication every two months. About work in the 21st century was one of the magazines, the other about democracy. It's great when a bunch of different experts speak on one topic. Quartz still like it.
Yevgeny Volnov:I love all western hipsters. Everyone whom Roma called, and I read them. In Russia, of course, I read Habr. I really like Medusa: here I am a rather involved, regular, active reader. I read RBC and, of course, sports.ru. Something is probably forgotten, but all of the above is included in my daily consumption. These are media that make a good product.
Sergey Baryshnikov: I read Pikabu, Sports, Habr. In fact, I am a very bad example: I consume content from the media at work, and not on my own will. It certainly leaves some kind of imprint. For fan, I consume news exclusively from Twitter, Facebook, Telegram.

- Today I heard several times that now in almost any industry there is an opportunity to do something new. If a person wants to write on a specific topic, for example, about sports, advise him to do something of his own, some kind of conditional blog, or go to a large aggregator, where he could theoretically have traffic - sports.ru, Tribune? The second question: which media areas are still not well covered?
Maxim Khryashchev: First you need to check how good you are the author. And is the topic that you want to write about really interesting to someone? To do this, it is reasonable to use the aggregator, to conduct a field study: write, look - read or not. If everything is very, very good, then you can start your own playground.
Yevgeny Volnov:I would suggest you do your own right away. So you will learn many things that you will not learn on another platform. Perhaps, at first you will write badly and uninterestingly, but someday you will learn. Instant success with you in any case will not be. I'm writing in my free time. Although I write very poorly, I have to work a lot on myself. On the second question - just last night I went to sports.ru to read about table tennis and was disappointed. All the media just terribly talk about this sport. There are only a few good publics on Instagram and Facebook. This is a very sexy sport, but not a single normal person would say so - I do not understand why. There so much can be revealed! There are several other categories on Sports that are poorly disclosed. Even the ski theme is presented so-so. And the list of topics can be continued for a long time.
Sergey Baryshnikov: I will answer the first part of the question. Eugene earlier gave excellent advice: do not listen to anyone. Therefore, go to the largest platforms, and where you feel comfortable, work there. Do not push yourself, do not be guided by the principle "everyone went to Instagram - and I will go there." If you are not comfortable there, then why stay there? Nothing good will come of it.
Anton Polyakov: You just need to love your business, which you do. Then many questions will disappear.
Roman Khudonogov: I would like to finish on a positive note: guys, make media! There are a lot of things said, but ... Go to Instagram, go anywhere - it's cool everywhere. I always have the desire to read something interesting.
In conclusion, we want to thank our sponsors and partners:
PS Reports from habrakonferentsii " Content " are available in video format . 16 hours of useful information, 26 speeches and 3 discussions on how to make cool content about technology and using technology. Promo code for discount: habr_online_promo.

The article is based on the decoding of the round table from our November conference " Content 2018 ". Speakers - Maxim Khryashchev ( Pikabu.ru ), Sergey Baryshnikov ( Bigpikcha.ru ), Evgeny Volnov ( Cyber.Sports.ru ) and Roman Hudonogov ( Setka.io ) shared their experiences and gave a number of valuable tips to budding authors and journalists. Does it make sense to open your own media - read under the cut.
Media Blockchain: Cryptomonettes for Content
- Several years ago, media based on the blockchain appeared in which the monetization process was implemented: people get tokens for posting, likes, etc. How relevant is this? Is there such a future for such media?
Maxim Khryashchev: For content to be of high quality, the user must have intrinsic motivation. It seems to me that for money it is difficult to write a good post. Another point that I noticed, after watching a bit for Voice (Golos.io): the content is influenced by the cost of cryptocurrency. When it grows, users are happy, they have many posts on this topic. The course falls - all upset, which is reflected in the content. I think this is not very interesting to a wide audience.
Sergey Baryshnikov:I do not see in the market successful implementation examples with tokens, although many platforms are experimenting with this. Once upon a time, LJ introduced its tokens, now Yandex.Dzen is trying to pay the authors money. Wins the one who first set up this work and will be able to offer a really tangible income to its authors. But when this happens is unknown.
Yevgeny Volnov: A super-perspective undertaking, but only as a research project. If you read the "Voice", then you will immediately understand everything. People discuss the whole blockchain mince, how to scale it ... The process is like in some kind of aquarium: you can see, laugh somewhere, and be glad that we are wrong.
Roman Khudonogov:Of these blockchain pieces, I liked Brave browser the most, although this is a slightly different approach. There, the user can pay for content that he likes. When you install the browser, you get 10-15 BAT on your account - this is like the currency of Thailand, only crypto. The more often they visit sites, the more they earn. If a user more often (compared to other resources) visits, for example, on sports.ru, then most of his cryptobates go there, to Zhenya. This model is more interesting, it motivates authors and publishers to make really interesting content. And a good scheme for the distribution of financial resources.
Recommender systems: should alcohol be recommended for an alcoholic?
- Let's talk about systems that advise additional content to articles. There is such a service as relap.io, which allows you to make personalized recommendations for users, taking into account their interests and some statistics. Another example is Echobox, which positions itself as an intelligent online posting. Question: how advanced are recommender services in general, and how advanced are recommender systems on your sites?
Maxim Khryashchev:We tried one recommendation system, I will not call it. As a result, she began to send users the most CTR-th records. But CTR does not guarantee quality, and we received a lot of negative feedback. Often there were very funny headlines, which I am even ashamed to voice. We have abandoned this system, but we use our own. We have the content shown in the tape, and the user can interact with him - like, vote for him, comment, etc. So it is easier and more understandable to him than with a third-party recommendation system.
When using such services, you need to understand why you are doing this, what goals you are pursuing. And assess whether it really helps somehow: it increases the viewing depth, affects your income or other indicators.
Sergey Baryshnikov:If the media has the resources to develop its own analytics services and recommendation systems, then no external system can compare with them. I respect the guys from Relap, we tried several times to work with them - but still, our internal recommendation system showed the best results, and we stayed with it.
Yevgeny Volnov:You need to use these systems to make money and put your own. As we did and it looks like everyone else. I also respect the guys from Relap, and all the guys who do something with their hands. But the recommendation systems remained, it seems to me, somewhere in 2013-2014 and did not move forward. You can do so much in this field, and no one does. Although these are obvious things. If there is someone who is engaged in recommendatory systems, come to me: I will tell you how we all can make money.
Roman Khudonogov:I also respect everyone, this is important. We had a different experience: first we made our own system, then we tried someone else’s. In the end, we realized that ours was still better. I am talking about Look At Media projects. In general, the idea itself seems a bit flawed - to recommend to readers simply abstract posts that he is most likely to click. The media has many different tasks, including information, education, and entertainment. But recommender services now perform only the last task, ignoring the educational function — in my opinion, extremely important for the media. If you read about some phenomenon - political, sporting, then it would be great to go over the recommendations to similar material. So that you, for example, dive into the topic, understand or discover something new in this direction.echo chamber , when you are in some kind of vacuum. And modern recommender systems are dragging us more and more there. And diversity is awesome.
- How do you think, will artificial intelligence come soon into recommendation systems?
Roman Khudonogov: There is a cool post on Facebook: if you see ML on a slide, then it is written in Python, if you see AI - in PowerPoint. That is, AI - this is too abstract, I guess. And the logistic regression written in Excel is quite ML and quite AI. So the recommendations already have ML, AI and everything else. It depends on how you prefer to call it.
Yevgeny Volnov:Recommender systems have two tasks: to narrow the person’s horizons or expand it. It seems that all this time we have been going to narrow it down. It seems to me that in the 2019-20-ies these recommendatory things will start to leave the agenda. People need to show some other things, and not just what they are likely to click on. Agree, it is strange to recommend a person alcohol, if he already drinks alcohol. Perhaps it is better to show him that there are many other pleasures in life?
Roman Khudonogov: Most likely, so soon they will not leave. The example of Chinese newsfeed Toutiao, which has 72 minutes per person per day with a 200-million active audience, confirms that business is moving in this direction. So the alcoholic will still be shown alcohol.
Maxim Khryashchev:Zhenya proposes to broaden the mind of readers. But the “or / and” recommendation system is very difficult to do: it does not yet understand how to show something new in order for the user to like it.

From left to right: Anton Polyakov, moderator of the round table, Habr's studio content manager, Pikabu publisher Maxim Khryashchev, Bigpikcha publisher Sergei Baryshnikov, publisher of sports.ru Yevgeny Volnov, and co-founder of Setka service (Look At Media) Roman Hudonogov.
About formats for presenting material and new branded media.
- What formats of presenting materials are interesting and relevant today? Does something new, promising, for which people cling?
Maxim Khryashchev: I did not actively study this question, but it seems to me that the main thing is in the essence of the content, and not in its packaging. In my opinion, the format in this sense is secondary.
Sergey Baryshnikov: I do not see anything new. The problem with the media right now is that nothing revolutionary is happening. A certain fatigue of media advertisers is connected with this. Therefore, they periodically go to the brand-media, to experiment with bloggers and so on.
Yevgeny Volnov:You can recall the Yandex narrative, although it is a weak example: a good product, but it’s difficult to call it a trend ... I think we don’t know how to entertain people, speak some complicated topics in simple language - but life becomes harder and harder. Moreover, we have a very suitable country for this. I like when the media can talk about complex things in the format of quizzes, but at the same time do not slip into some kind of tin. We are on sports.ru, I am constantly talking about such a project as ranker.com: look, drop dead model, very simple. I do not remember other examples when the content would be repackaged into some kind of interactive, which a person can perceive in half a minute.
Roman Khudonogov:I found the answer while driving here in a taxi: a study by Reuters and the University of Oxford, which is called Time to Step away from the 'bright, shiny things'? The guys interviewed 39 leading journalists from 17 countries on the topic “what to do next”. It turned out that for the last five years, there have been throwings from trend to trend. Type: “Oh, storiz! Let's make a story, it's very cool! .. ”“ Quartz launched on Facebook Messenger ”- you sit there, suffer, read the news in the messenger, you don't understand why - but it's cool! Conclusion - it would be good to return such formats as journalistic investigation, narrative, and concern for the reader. This is cool, in my opinion. In general, when there is a trend, a wave of advanced technology, then at first everything rises, then - the bottom. After all suffer, and then everything starts to grow again. Here we are moving along such a J curve funnel.
Now it’s more interesting for me not so much the format as the specific publisher. The fact is that recently there are interesting alternatives to the classic media. By classic, I mean online media. Books and magazines are already somewhere far away. Now brands began to create more and more incredibly cool media. These publications are outside the model “let's accumulate an audience and sell its attention to advertisers”. They perfectly reveal topics, especially professional ones. Now I don’t know a single traditional media in which I could read many cool articles on a topic that interests me. But it seems to me that over the next 5-10 years we will see a fundamentally new media landscape, where the brand will stand on a par with the publisher. He will also become a publisher and will often solve the reader’s problem better. The fact, that brands are manically focused on User eXperience, they cannot just say: “Buy, please, from me! I'm cool! ”They need to tell the story so that the person also associates this cool message with the brand.
For example, the Intercom brand is such a thing that appears on the right of every third site and is terribly annoying - publishes a cool magazine Inside Intercom. They write not about how cool they are, but how they make the product, how they organize the support service. This is a super cool trend! The second example: I do not know any media that talks about extreme sports cooler than Red Bull. And if you ask the audience what kind of media is the best in Russia, it tells about money, how to save money - most likely, the majority will call the publication of a brand that we all know. And this is really cool. It is possible that some brand will soon release a political publication. It seems incredible, but I think it will.
Distribution and new distribution channels
- What new distribution channels have appeared lately? What did you experiment with? What channels have you gone well or caused interest?
Maxim Khryashchev: Experimented with messengers, tried Weiber and Telegram. The same groups appeared in Vaibers as in the Telegraph. They differ from old ones in that they are displayed among the list of your personal contacts. But still we did not go. It turns out well with Telegram: they scored 164 thousand subscribers, advertising is sold in the group. After some time, we may pay back investments in the development of the channel. Very well monetized channel in VKontakte. But Facebook does not pay off, but we have historically not developed a relationship with him.
Sergey Baryshnikov:In 2017/18, many media outlets hooked on the Yandex.DZEN traffic needle. Probably, everyone will be hurt in 2019 to jump off of it, as it was four years ago with Rambler. Yandex.Dzen, in my opinion, is now “number one” by the way it pours traffic. And then further - who has learned to work with him, he can monetize it.
Yevgeny Volnov:The word "distribution" has crammed. I hope in 2019 the frequency of its use will fall. So far, nothing new has appeared - and thank God: we would have to hire people, explore, stuff cones. And we are already tired of this. Everything is changing rapidly, someone constantly sets experiments on us. They are somewhere successful, somewhere not. What to do, you have to live with it. Although everything has turned into some kind of bytovuha, which is just boring to do. I agree with Roma about the media: in the media it becomes crowded that someone comes to you - advertisers, platforms and so on - and all the time tells how to do it. And brands do not have these restrictions. There are other frames, but they allow you to do something that classic media simply cannot afford because of some patterns. Media turned into brick factories: Here is such a distribution channel, let's ... The question of distribution is part of this factory for the production of an absolutely standard, uninteresting product. This is terrible, we must move away from it.
Roman Khudonogov:We began to forget that your website and your application is incredibly cool. You have direct access to your entire audience at once. And on Facebook you accumulate a group of 150 thousand, publish a post and have access to 2% of your audience. Distribution needs to be done, otherwise you will lose. But own content hubs in the form of their applications and sites - that's what's really cool! You can collect data, recommend and understand. No one can do it better than you. And the most important thing is that you are within arm's length of this very audience. For example, you can send an email to contact with the user. You do not need to rely on the mechanism of distribution of attention, which uses a particular service. I wanted to post a message, and they say to you there: “You can forget it in your group for money.
Yevgeny Volnov: There are about a hundred of our own mobile applications in different languages in our company. I know how much we spent on them and how much we earned, so I want to tell you: do not make mobile apps, especially if your media is not sports.
Roman Khudonogov: He specifically says so, I assure you: “Guys, please do not go to any platforms! Do not do the media, it will end badly ... ”

- How much did the reader's attention spread through these channels?
Maxim Khryashchev:We had this story with one of the development engineers. When he met his girlfriend and talked about his work, she asked: “Are you working in a group on VKontakte?” That is, she didn’t know about our site in principle. So due to additional sources of traffic, we acquire a new audience, which does not know our original sites, sites, applications.
Is it possible to measure the quality of content
- With the massive advent of bloggers, the quality of content has sunk. Is there a demand for a better product? And how good can user-generated content (UGC) be?
Maxim Khryashchev: When we started, the quality of the content was the most important factor for me. I judged by myself: why I will use the resource? And I decided that I need to go every day and see quality content. Somewhere blog content and sags in quality, but still there are some very cool materials. I like the same VKontakte experimenting with the Longrid. In general, it is very difficult now to offer content to the user, and the main struggle is underway for quality. Therefore, it seems to me that, in absolute figures, the volume of high-quality content, on the contrary, increased.
Users can make good content, but you need to understand their motivation: it definitely will not do something worth 10 rubles - or how much you can earn there in Golos. With the right motivation, the quality of UGC will be. Especially if by quality we mean not the design, but the essence of the content.
Sergey Baryshnikov: Content quality is a subjective value. I will argue about the fact that bloggers have reduced the quality, on the contrary: take the same Dudya, who raised the bar on YouTube for everyone and for the media as well.
- Yes, there are good examples. But generally, on YouTube, the quality of blogging content is very controversial.
Roman Khudonogov:The quality has not deteriorated, just more channels, communications, people. Take the millionaire channels: there were, for example, 100 bloggers - it was 12,000. Probably, some integral quality has decreased, but in general, everything remained at the level.
Yevgeny Volnov:I don’t like this content option like quality. Let's draw a parallel with art: we do not argue in the gallery about whether a quality or low-quality picture. It can be boring, uninteresting, with lots of borrowing and so on. But I can not say anything about its quality. This is in any case some kind of statement. Even a single letter is a statement that may be interesting in context. And if the bad texture, but the text of 20 thousand characters - who will read this? It can be cool and high-quality analytical content that no one needs. Because the author does not understand what people need, what is now important, how to convey your emotions. I am for any content, therefore I have no complaints about UGC, any comments, etc. The main thing is that everything should be done with talent. Then it's cool.
Roman Khudonogov: I believe in quality content - in terms of the existence of such a term.
It seems to me the best thing is when art is combined with craft, that is, with production and class formats.
Again, returning to brands and to a request for quality content: we want to give our readers maximum benefit. I really like it when the guys do the work for the reader, when they try to figure out a topic. So that you do not need to look for any other sources, and when this post is really cool decorated, when you are pleased to consider and read it. I would call it quality content, and it, in my subjective opinion, can defeat the mentioned cloths of the text, not very interestingly written, and so on. If we go back to some retro media, then there is just a book on an interesting topic - and a cool book written about the same. The second one wins a lot, so high-quality content is awesome. I personally have a request, I want high-quality content.
Maxim Khryashchev:I really liked Zhenya's analogy with art. Indeed, by what parameters do you like the picture, how would you rate it? Probably, this is the author’s style, the plot itself, the size - it can be all over the wall or small ... Similarly, you can evaluate the content: how it is designed, how beautiful the photos are, whether it is easy to read, whether there is humor, how facts are picked up, and so on. That is, there are a number of parameters, and it is already possible to judge them collectively. It turns out that there is a certain scoring system inside of us, according to which we evaluate how much we like or dislike this content.
FactChecking: Benefit or Harm
- We know that people can be wrong. Sometimes posts that have a fatal error collect a million views and hundreds of likes. And no one will pay attention, if somewhere in the 15th comment someone writes: “Guys, things are not really so ...” Do we need any kind of billing system? And in general, how UGC can be reliable?
Maxim Khryashchev:Yes, there is such a problem. Some users post unreliable content, others refute it, write refutation or require, for example, to delete a post. On the other hand, in a publication without a UGC, there can be an author who will write inaccurate content - and no one will give him a feedback that this content is unreliable. We work with this on the basis of feedback of users. If they show us, the support service, some facts that refute the post, then we delve into it, study it. According to the results of the check, we either remove the disputed content from the site into a specific section that no one sees, or we edit it, or do something else with it. Users have a serious demand for fact-tracking. This is a very important indicator of content for you to read, and this is true.
Sergey Baryshnikov:As a consumer and, on the other hand, as a person who is close enough to money in a company, I consider factual billing to be an absolutely harmful thing. This is some kind of lobbying, manipulation, etc. First, it can only look like an ecosystem solution: different players sat down and agreed on some general rules like: “We need to improve the ecology of our information environment, let's invest a couple of tens of millions of rubles and our brains to solve this a problem for society. " Can you imagine such a situation in Russia? Me not. Secondly, factchkking is a good way of manipulation, including for some regulatory bodies. That is, to any media that has stumbled somewhere, they can come and say: “Guys, we’ve spoiled you, you’ve become some bad, now we will punish you ...” We see similar examples. Who speaks about factbooking at all, how about some super-big problem? Facebook and similar players are practically monopolists. Not sure they want to bring honesty into our lives. In addition, our system is still capable of self-regulation. The same Navalny is an excellent fact-tracking system: fun, interesting. I think these things are enough. Plus it is technically very difficult to do this.
Roman Khudonogov: In my opinion, this is a question like personal hygiene. I wash my hands, but to go out and shout about it is not very cool. I agree with the guys that a good solution is factbooking, based on some kind of feedback loop: users complained, you checked.
Yevgeny Volnov: Yes, there may be some kind of hybrid stuff here. At the same time, we, the media with large reach, need to think about how to teach people some kind of consumer culture, to explain many things to them. And factual billing should be created inside. There is definitely no one right way; it can happen differently for everyone.
Robots - structure, man - creativity
- When articles for journalists will begin to write artificial intelligence machine? How realistic is this?
Maxim Khryashchev: I think these articles will not be very interesting. It is necessary to invest some kind of individuality in the content so that other users with a different personality would be interested in reading it. So I do not really believe in it.
Sergey Baryshnikov: On the contrary, in a year or two we will see simple materials, news from some kind of automatic media. Probably, they will arise in our field. But how deep they step further - no one knows.
Yevgeny Volnov:Some kind of hybrid models will definitely appear. Artificial intelligence can help a person put the right word, determine which of the two synonyms is more clickable, or choose the most clickable from several headers, send a post to Facebook or Yandex.DZen.
Roman Khudonogov:During the last Olympics, the robot company Byte Dance, which makes newsfeed Toutiao, composed and wrote news on sports topics. They were well structured, understandable, and not easy to say. To create each text, it was necessary to look at several sources, to understand who beat whom, to make a small podochka. There were a lot of these kind of publications, and it really worked. That is, in formats where the structure is clearly visible, robots are already writing. That's right, let the journalists do other things - more creative.
Its media: do not even try or not listen to anyone
- What advice can be given to editors and authors? What technologies and services - from new - would be useful for them? And in general, what is the main thing in the content?
-Evgeniy Volnov: Do not pay attention to anyone! Do not listen to anyone. Formats? Hammer! Examples, including Dudya, with whom I worked in the same company, sat in one meeting, saying that one should not drive oneself into the framework. Yura never tried to do something like “this is to make money ... to say this and this ... and to make it look expensive.” It would be trite and uninteresting. In our time, when so many opportunities - both technological and in terms of topics - you just have to go and do.
Roman Khudonogov:I am presenting Setka service here, so recommending it to someone is not very cool. But still I want to say that design is a very important thing. If you have created a certain story, article or something else, do not forget: the correct packaging in the form of a cool structuring of the text and everything else - the work done for the reader. An important thing that is often missed, because of this, we organized our small enterprise: working in the Look At Media publishing house, we came to the conclusion that you can tell stories more interesting, better. We are trying to change this industry a little bit.
Sergey Baryshnikov: I have literally the last advice. Do not try to open your media, because there is no money there.
Yevgeny Volnov: Maybe there is no money. But then there is an opportunity to make a real revolution. Spit on the money, go there and do super cool things! Now there is weak competition, there are few who do very cool things. This field is free.

How to generate feedback, where to go to the beginning author and other questions of conference participants
- At the recent RIW, the media unanimously criticized Yandex.DZen for not giving them enough traffic. Today you mentioned that Zen is paying authors money. How exactly? Now, there seems to be only tapes of the coolest Russian media.
Sergey Baryshnikov: Each author, upon reaching a certain number of subsidiaries - seven thousand or ten - can insert Yandex.Direct blocks and receive some money. Major authors receive, in my opinion, several tens of thousands of rubles a month.
- Recently watched Habr's admin panel. On average, a post with 50 thousand views has 100-150 reactions (plus or minus karma). What are the new ways to involve users in feedback - so that they press something, express their opinions? For example, Medium has claps, you can pat ...
Roman Khudonogov: It seems to me that the reaction is an atavism. If you can understand the behavior of the reader, like it or not, why ask to press it on claps? Therefore, most likely, they have become smaller.
Yevgeny Volnov:When we force a person to explicitly respond to the content, he becomes an accomplice. This is not the way to know what he thinks - as if it were important to us. But this is a way to involve him a little bit in content creation. I like the widget from Disqus, the comment service: you can read the article and display your mood as you know - you know how in Sberbank, how you were served. It may seem strange, but it's fun. Plus, all sorts of things like “You read it - and what do you think?”, “Who is right, this one or the other?” I tell about the fact that we entertain poorly. Only in this direction there is still a lot to be done, and no one does anything. By the way, I do not understand why.
Roman Khudonogov:In UGC projects, this resonates: it is cool to involve a person and give him not even an illusion, but the ability to somehow influence. This is cool, I agree.
Maxim Khryashchev: The guys said about the user who leaves a reaction. But it is equally important to think about the author of the content, who wants to get some comments, reactions, advantages. It is important to assess the audience - well written article or bad. Regarding statistics, it is important to clarify how many of the 50 thousand who viewed were registered users, how many logged in. After all, not the entire audience can leave this reaction.
Yevgeny Volnov:By the way, we talk so much about all sorts of donations to the authors. Does anyone know a tool that could be easily embedded on the site and that would allow without any hemorrhoids to directly pay the author, so that it would be some kind of scalable system? I dont know.
- What information resources do you personally read? Western examples are especially interesting.
Roman Khudonogov: I really like the MIT Technology review, especially their electronic journal. They began to publish the publication every two months. About work in the 21st century was one of the magazines, the other about democracy. It's great when a bunch of different experts speak on one topic. Quartz still like it.
Yevgeny Volnov:I love all western hipsters. Everyone whom Roma called, and I read them. In Russia, of course, I read Habr. I really like Medusa: here I am a rather involved, regular, active reader. I read RBC and, of course, sports.ru. Something is probably forgotten, but all of the above is included in my daily consumption. These are media that make a good product.
Sergey Baryshnikov: I read Pikabu, Sports, Habr. In fact, I am a very bad example: I consume content from the media at work, and not on my own will. It certainly leaves some kind of imprint. For fan, I consume news exclusively from Twitter, Facebook, Telegram.

- Today I heard several times that now in almost any industry there is an opportunity to do something new. If a person wants to write on a specific topic, for example, about sports, advise him to do something of his own, some kind of conditional blog, or go to a large aggregator, where he could theoretically have traffic - sports.ru, Tribune? The second question: which media areas are still not well covered?
Maxim Khryashchev: First you need to check how good you are the author. And is the topic that you want to write about really interesting to someone? To do this, it is reasonable to use the aggregator, to conduct a field study: write, look - read or not. If everything is very, very good, then you can start your own playground.
Yevgeny Volnov:I would suggest you do your own right away. So you will learn many things that you will not learn on another platform. Perhaps, at first you will write badly and uninterestingly, but someday you will learn. Instant success with you in any case will not be. I'm writing in my free time. Although I write very poorly, I have to work a lot on myself. On the second question - just last night I went to sports.ru to read about table tennis and was disappointed. All the media just terribly talk about this sport. There are only a few good publics on Instagram and Facebook. This is a very sexy sport, but not a single normal person would say so - I do not understand why. There so much can be revealed! There are several other categories on Sports that are poorly disclosed. Even the ski theme is presented so-so. And the list of topics can be continued for a long time.
Sergey Baryshnikov: I will answer the first part of the question. Eugene earlier gave excellent advice: do not listen to anyone. Therefore, go to the largest platforms, and where you feel comfortable, work there. Do not push yourself, do not be guided by the principle "everyone went to Instagram - and I will go there." If you are not comfortable there, then why stay there? Nothing good will come of it.
Anton Polyakov: You just need to love your business, which you do. Then many questions will disappear.
Roman Khudonogov: I would like to finish on a positive note: guys, make media! There are a lot of things said, but ... Go to Instagram, go anywhere - it's cool everywhere. I always have the desire to read something interesting.
In conclusion, we want to thank our sponsors and partners:
- mega- sponsor RuVDS
- water partner "Borjomi"
- of the Letters Metter partner
- conference of Oleg Bunin
- video partner FU2RE
- information partners Adindex, Pressfeed and Media2
PS Reports from habrakonferentsii " Content " are available in video format . 16 hours of useful information, 26 speeches and 3 discussions on how to make cool content about technology and using technology. Promo code for discount: habr_online_promo.