"Illegal and toxic content in advertising has practically disappeared"
Four years ago, the law on mandatory labeling of online advertising was passed. Vladimir Logunov, Deputy head of Roskomnadzor, told Izvestia about the fruits that the implementation of this initiative has already brought and what to expect in the future.
— Tell us why the labeling of online advertising was invented, what tasks were set?
— The idea appeared in early 2020 or even a little earlier. By this time, the labeling of goods "Honest mark" had already been introduced. The task was set: starting from the same goals as the usual labeling of goods, to ensure the traceability of advertising placed on the Internet. However, unlike ordinary goods and even ordinary advertising on billboards and on television, online advertising has one feature. This is her fleeting presence on the show grounds. Right now there's one banner, and seconds later there's another. There was no record of the impressions and their number, and there was no reliable information about the cost of these impressions. And the initial purpose of labeling is to confirm the validity and reality of spending on advertising on the Web, to increase the share of the white market. This means making it transparent and increasing tax revenues. After all, the cost of online advertising was often only on paper, but in fact it might not be at all. The tax base is immediately significantly smaller, and the money could simply be "withdrawn". And in some cases, direct advertising customers did not fully understand what they had paid for and why customers did not come to them.
Another goal is to comply with the requirements for advertising in terms of the Federal Antimonopoly Service. There were complaints from market participants about violations of advertising laws, but the content was no longer to be found. And, of course, the requirements for Roskomnadzor's content, so that it does not contain information prohibited by the legislation of the Russian Federation, the appearance of which we monitor on all sites.
Finally, we had a request from the market to create a single analytics tool that would provide online information about the objective state of economic indicators — market volume, growth, assessment of advertising trends, and popular promotion platforms.
— Why did Roskomnadzor become the operator of such a system?
— Our agency is an active advocate of state policy in the field of traditional values and oversight of the content of everything that is broadcast on the Internet, including advertising, social advertising and other types of information. Online advertising itself - to paraphrase a well—known phrase by the founder of Russia's first advertising agency — has become the engine of meaning. Banners next to the article you are reading work like the 25th frame, but only more efficiently. Unlike TV, advertising "engines" see what kind of person is on the other side of the screen, and they know exactly what they will be interested in. In the early 90s, TV commercials also carried a lot of foreign culture and meanings that were not always related to our values, but they managed to regulate this. That's not the case on the Internet yet.
Another argument in favor of labeling online advertising is the need to introduce economic measures against violators of information legislation instead of blocking resources that are important to users. And for this, we need both a regulatory framework and a technical tool in one hand.
— How did the changed geopolitical situation affect this work?
— In the context of the information war unleashed against Russia, these tasks have become even more obvious and require solutions. Through advertising networks, they began to deliver content and meanings directed against Russia, including prohibited information. In addition, advertising has become a source of financing for banned sites, foreign agents, and monetization of extremist activities. Advertising networks have begun distributing links to phishing resources and the resources of scammers whose purpose is to deceive citizens. Budgets even from pro-government companies could end up on websites of banned resources or sites with an anti-Russian position.
It was important to create a solution that would help stop the dissemination of this information and at least reduce the number of sources for financing illegal activities. In order to achieve these goals, it was necessary, in addition to the regulation introduced in stages, to create an entire digital platform that would become a technical tool for recording and controlling all information about distributed advertising and ensure that the necessary labeling was provided to confirm that the data on this advertisement was entered into the platform.
— What do I need to get marked?
— To receive the marking (a special digital code, or ERID), it is necessary to register the creative (the advertisement itself), specify the advertiser's data, region, INN, OKVED, and contract data. Transmit the content of the ad (video, image, text, audio). The ERID must be specified in the advertising message or next to it (there are many different ways to do this, including native ones). After advertising is distributed, provide reports (for example, information in the act of completed work), information about all distribution sites and the volume of impressions, as well as their cost.
All information is included in the unified Register of Online advertising (ERIR) created by us. As a result, the entire advertising distribution chain appears in it: from the advertiser to the final distributor, the content itself, the content of the advertisement, its cost and the sites where it was shown.
— But this is a huge amount of data for both the agency and market participants. How did you technically manage to do everything by the time the law came into force on September 1, 2022?
— Initially, it was assumed that all market participants would directly transfer data to the registry. However, in 2021, a pilot was conducted with the participation of the largest advertising systems, after which it became clear that a huge and expensive storage would be required to store the entire amount of information. Therefore, we decided to create specially selected advertising data operators (ORDS), which have become technological intermediaries between market participants and ERIR, and are engaged in receiving and storing advertising creatives — images, texts, audio and video clips. By implementing this architecture, we were able to significantly reduce the load on the registry — only links to these creatives and metadata are transferred to it.
In addition, the services for issuing ERIDS and transferring information to the unified register of online advertising have been integrated into major Russian advertising systems, providing almost complete automation of the labeling and information transfer process. This made it possible to make the online advertising labeling mechanism almost "seamless" and launch advertising campaigns as quickly as it was before.
The ARDS themselves were established by the largest IT and advertising companies, and were accredited by Roskomnadzor. The involvement of the largest Russian companies in the implementation of the project ensured a customer-centric approach to accounting for information, as the ORDS have implemented and are refining their systems taking into account the wishes of their customers. Someone is more focused on bloggers and provides them with the functionality of automatic reporting on impressions. Some are aimed at large and medium—sized businesses and provide, first of all, the opportunity for flexible and rapid integration of information systems of market participants with ORD systems. As a result, market participants have a choice among seven high-tech companies that differ not only in names, but also in architecture, interfaces, and customer service.
— And what is the COMMANDER himself? What is its architecture?
— It is a digital platform with a huge database. The system is a completely Russian development and has no analogues in the world. The software used to build the ERIR is specially developed for the registry by Russian companies. Today, the registry contains information on more than 8 billion advertisements, statistics on trillions of ad impressions on various platforms (websites, social networks, messengers, etc.) and more than 1.5 million sites. The EPIR takes into account the advertising of 1.29 million advertisers each year. Last year, their advertising budget, reflected in the EPIR, amounted to 858.6 billion rubles.
Advertisers themselves have become users of ERIR. They can check on which sites and to what extent their ads have been distributed (and most importantly, that they actually were). For citizens today, the registry is like a quality mark, they can check and make sure that there is a very specific person behind the advertised product or service who is responsible for the information provided in the advertisement. Anyone can check the labeling of ads in the EPIR by logging in with their Gosuslug account.
I can confidently say that ERIR is probably one of the largest government information resources. The system was designed in such a way as to maximize the linear growth of information without changing the architectural scheme. In addition, the registry is not tied to specific hardware, which allows us to be flexible in the current market conditions. The system uses the most modern technologies available to developers on the market: microservice architecture, streaming distributed analytical database, and other features that are also used by leading Russian Internet companies. The registry has created personal accounts and dashboards for government agencies. The Federal Tax Service, FAS, and RCN receive data through them and the API. In total, we serve up to 100,000 simultaneous sessions without sacrificing speed.
— But there were difficulties in developing and creating the registry, weren't there?
- yes. At first, the architecture of traditional databases was used, which made it possible to quickly collect information from September 1, 2022. However, a month later, it turned out that the chosen architecture could not cope with the huge amount of data. The database quickly accepted the information, but did not allow for its prompt reconciliation and analysis. We decided to urgently replace the database with ClickHouse, which was better suited for processing large amounts of data and conducting analytical cross-sections.
At the same time, we were working to support APIs (interfaces for advertising data operators) through which data was transmitted to ERIR. As a result, in November 2022, all the collected creatives were successfully uploaded to a new database and their relationship with related entities was checked: contracts, additional agreements, acts and counterparties. With its help, we quickly uploaded billions of collected creatives and ensured uninterrupted reception of new ones. And most importantly, by December 31, we have redesigned all the ERIR interfaces and, above all, the personal accounts of all market participants. Just a month after the architecture change, any advertiser could already log into the registry and view the information. We completed all this work in less than a month. Which is an extraordinary result for such projects.
Overall, we have come a long way and achieved our goals. The obligation to disclose all information through the EPIR, to put a token, is very disciplining for players. Illegal and toxic content in advertising has practically disappeared. It is marked even for foreign social networks, because the advertiser is interested in this.
— And how did you set up work with the quality of the data of the hordes?
— At first, we did not impose strong restrictions in order to start collecting information. But at each step we add more and more general reference books, format and logical controls that do not allow us to accept dirty and junk data. In addition, there are integrity checks in the system itself. For example, if there is a link to a registered agreement when registering a creative (creatives must be registered to an agreement with the advertiser), we will accept the data. If there is no such agreement, we will not accept the creative and will send an error message. To get feedback, since there is a lot of data, we have created a data quality confirmation service so that the ORD can check whether the information has been accepted into the EPIR or not, and clarify the current status of an entity. At the same time, we conduct regular checks on analytics with various associations, including ACAR, and the market, study anomalies and identify errors so that the data is more accurate.
Three percent deductions have now been introduced for income from the distribution of online advertising. Monitoring the reliability of the data is now even more important, as errors will be costly for market participants. Additionally, the accuracy of information can be improved with the introduction of a kind of counter, a single technical means of objectively recording impressions and clicks of online advertising. This is still in the plans, but the topic has already been raised by the industry itself at SPIEF.
— By the way, tell us about the introduced three percent advertising fee, what purpose does it pursue, besides replenishing the treasury?
— The amount of deductions is calculated in the unified register of online advertising based on data that market participants have transmitted through the ORD to the ERIR. This information reflects the results of the provision of advertising distribution services, and it is the basis for determining the amount of the fee. Since deductions and reductions in the tax base are calculated on the basis of data provided by market participants themselves, they are interested in accurately reflecting the information in the EPIR. This should reduce the use of gray schemes when some payments are not included in official reports.
Deductions work as a natural filter. Bona fide market participants will structure their processes, while unscrupulous ones will either adapt to the new conditions or leave the market. Funds in the form of mandatory contributions received from the sale of online advertising distribution services will be used to support domestic online platforms that participate in government projects. This is done by analogy with the collection from telecom operators to eliminate digital inequality. It is fair that it is necessary not only to maintain access to the Internet, but also to develop domestic platforms in it. And participants in the Internet platform market have not yet paid anything, unlike telecom operators. Therefore, in fact, the law has equalized the responsibilities of all those who represent the Internet industry in order to support the sovereign information space.
— What is the mechanics of collecting deductions? Which members of the chain pay it?
— According to the draft government decree, the fee will be charged only for the first act of completed work in the advertising distribution chain. If an agency agreement is concluded between the contractor and the advertiser, we will also charge a fee for the agency fee.
In order for the collection system to work correctly, we need to build chains, analyze millions of reports every day, and update information for our users in their personal account at ERIR. For example, we have received information about a new act of completed work. First, we need to integrate the act into the chain and understand which contracts have already been concluded and which accounting documents have been submitted. Based on this, the system concludes whether we should charge something for the income specified in this act or not. If they have to, the payment calculation base will change and the calculations will be updated. If one of the participants in the chain has corrected or deleted information, the system also recalculates it.
To do this, we update billing information daily. In addition, we need to consider exceptions for certain categories, including social advertising and media. To verify this, we also analyze millions of acts and conduct billing reconciliation — who falls under the three percent fee and who should be released.
In addition, ERIR has its own deduplication logic. There are cases when the customer and the contractor report on the results of the work performed by uploading the same reports. The system sees that, for example, the same work report has been uploaded twice (both the customer and the contractor) and will use only the latest uploaded version of the report to correctly calculate mandatory deductions. But if one of the parties to the agreement makes changes to the previously uploaded act, for example, assigns a different number to the document, these are already different acts for ERIR, both versions will be relevant and used in calculating deductions.
— But from the point of view of other information. After all, a lot of information is broadcast through advertising networks and is not an advertisement? Or, for example, paid information promotion campaigns. What are you planning to do with them?
— It is expected that the growth of the advertising market this year will be higher than last year. However, due to stricter criteria for classifying information as advertising, many of its participants may resort to a formal approach in order to avoid complying with the requirements of the law. As a result, the total cost of information and influencer marketing is likely to increase, but the share of what is officially classified by the state as advertising may decrease. In other words, it may happen that content uses the same inventory, advertising networks, and platforms, but its distribution becomes opaque and does not achieve its goals. which we've already talked about.
Everything in the white zone works according to the rules and, as we can see, really brings results. And it's not just transparency and traceability, payment of royalties, and so on. Advertisers appear in such an environment who, by distributing their commercial advertising, bring very valuable meanings there — this is love, family, patriotism, development and unity of our country. If any information on the Internet is supplemented with such content, it will solve a very important task of communicating to the Russian audience through powerful advertising tools the concepts of traditional values enshrined in the presidential decree.
Переведено сервисом «Яндекс Переводчик»
