Eight of Germany's 10 largest publishing houses are working to create a single database of their readers. In parallel, user data combines The Guardian, CNN, Financial Times, Reuters and The Economist.

1000 sites, including the largest tabloid Bild, began to exchange data sets. Axel Springer, Gruner + Jahr, Bertelsmann Group and Der Speigel postponed the traditional rivalry and share valuable information about their readers with each other (read - with direct competitors). So publishers respond to the actions of Google and Facebook: today, corporations control 85% of the digital advertising market , while the rest of the players have to share the rest.

“Our biggest competitors are not conservative publishers. Together we must join forces against Google, Amazon, Apple and Facebook, ”said Karsten Schwecke, CEO of the sales house Media Impact, owned by Axel Springer.
 

How did the idea come about

 
It’s very difficult to sell media advertising without analyzing big data. Yes, each publisher has a certain amount of user information in terms of targeting, segmentation, and behavioral factors. But today this is not enough. Advertiser requests are constantly growing. They want to reach the audience segment they need at the same time on multiple sites. A separate media outlet and even a publishing house do not have the necessary data for such a campaign. In other words: to make money in a market where advertising is targeted by segments, modern media needs big data. Moreover, not all media can boast of the presence of specialists with sufficient technical qualifications in the analysis of big data. Yes, and many simply cannot afford to maintain such a service.

Participants of the Pangaea Alliance project, created last year by the world's leading media companies: The Guardian, CNN International, Financial Times, Reuters and The Economist, also came to the need to combine their data for successful advertising targeting.

“Data is critical,” said Pangaea Project Manager , Guardian News & Media Revenue Director Tim Gentry. According to him, by exchanging primary information about their users, media companies create "unique and attractive segments of the audience." For example, subscription information from one publisher can be combined with behavioral information from another. As a result, a detailed user profile is created for which advertisers are willing to pay.
 

How the new database works in Germany

 
The raw information from the publishers goes to a single platform called Emetriq , which is managed by Europe's largest telecommunications company, Deutsche Telekom . The latter sifts and cleans the data to create narrowly targeted, high-quality segments of the audience, using which publishers will improve their advertising offers.

Each publisher pays a fixed amount for using data segments from € 4,000 ($ 5,000) to € 15,000 ($ 17,000). The price depends on the page impressions on the sites and the amount of data used from the general “bank” when displaying advertisements.

Such projects are not launched in Germany for the first time. In 2010, the country's seven largest publishing houses, including Axel Springer, created a joint venture, AdAudience, which sells advertising based on the Programmatic direct principle. This is how Emetriq gains access to publisher data, each of which remains anonymous. Publishing houses can use social and demographic information about users, data about their behavior on the Internet, purchases, as well as the semantic core and context of websites. This is how publishers create high-quality reader profiles and advertising segments that appeal to advertisers.

All segments created by Emetriq are reconciled with the consumer panels of the GfK research group.. Deutsche Telekom also adds some of its own data in exchange for being the only advertiser with access to it. Together, this gives an accuracy rate of 87%, says Daniel Neuhaus, CEO of Emetriq.

Segmented data is sent back to publishers who can select a segment to run through their own data management system (DMP).
 

Why share data

 
Combining data helps reduce publishers' dependence on Google and Facebook, which now dominate the market thanks in large part to big data. "Draining" resources into a single bank, German publishing houses create an alternative and relatively large data set that will allow them to earn more, Business Insider notes .

According to Daniel Neuhaus, German publishers are still far from the level of media giants, although they are becoming closer in quality and quantity of data: “No one suffers more than publishers and sales houses in Germany, because they have insufficient data and their arrays of unstructured information about users can never even compare closely with Google and Facebook. "

“The concept of Emetriq is to help people be innovative wherever data is needed, and to make it a reality,” said Neuhaus. According to him, the fact that Google, Facebook, Amazon and Apple collect data and store it at home, without sharing with anyone, "will ultimately kill innovation."

“The real benefit of the merger is to create better segments that customers really need, along with high reach,” said Stefan Kreutz, CEO of AdAudience Joint Venture. According to him, it is not so difficult to get a good quality advertising campaign with low coverage or, conversely, high coverage with poor quality. “But doing both at the same time is a difficult task,” added Stefan Kreutz.

Ultimately, the benefits to publishers are obvious. They get extended, anonymous data segments from the “common boiler”, which is much larger than their own brand portfolio. Thanks to this, publishers can increase their CPM (cost per thousand impressions of the ad unit). Since the project was launched recently, AdAudience does not give figures demonstrating the growth of CPM, but, according to Stefan Kreutz, there are already the first signs that it will be “significant”.
 

How data aggregation will change the market

 
Combine the data and many commercial brands engaged in direct sales. This allows them to target and retarget customers, to receive information about their purchasing interests and when they are ready to make purchases, as well as about gadgets and equipment that they have.

For example, at the end of last year, the UK affiliate network Skimlinks launched a “data cooperative” called Audiences by Skimlinks , which brings together more than 55,000 publishers and about 20,000 retailer sites. Based on information about user behavior, segments are created that allow you to successfully target ads.

Alicia Navarro, CEO and co-founder of Skimlinks, explained the benefits of participating in the project. For example, a user on the publisher’s website, while reading an article, clicks on a link about a woman’s bag and is redirected to the website of a store selling this product. If, as a result, a purchase is made, the following chain occurs: page view> click> conversion. Even if the user as a result does not buy anything, the publishers and retailers participating in the project receive a behavior model (pattern) for a specific user. When he appears on the website of another publisher, he can again be redirected to a page selling women's handbags, thanks to a cookie file that Skimlinks will leave on the user's computer. According to the creators of the project, it led to an “almost unbelievable” increase in conversion.

According to informationMarketing Land , advertisers pay for segmented data on a CPM basis. Income is distributed among the members of the “cooperative” on the basis of the data they provide: the more valuable the data, the greater the profit. For example, page impression information costs less than click information.

Alicia Navarro notes: “The more data we have, the more accurate we can be.” According to her, the project Audiences by Skimlink is also unique in that the participants of this “cooperative” see “all the way from content to a commercial transaction”, while other data providers, as a rule, have only retailer data or only customer data intentions.

Another project of this kind is successfully developing in the USA. New York-based Bomboramonitors search queries, document downloads, webinars, registration at exhibitions, viewing articles and blogs, video consumption, likes and shares on social networks and other evidence of activity of entrepreneurs who are looking for certain goods and products. This project involves Forbes, Aberdeen Group and about 2,500 other sites that provide data on more than a billion monthly interactions with their visitors - usually these are business buyers. Advertisers and agencies then use this information for marketing and sales, offering it based on targeting to interested business companies.

Bombora claims that it was their company that became "the first aggregated source of behavioral data for B2B, which created the" first of its kind "data cooperative of premium media companies."
 

Is media cooperation possible in Russia?

 
So far, no such hearings have been heard from Russian publishers. Experts note that the media environment in Russia is not yet ready for such projects. We can tell about the experience of working with big data in the media in the Russian market using our own example - Relap.io . More than 2,000 media on Runet use the Relap.io

recommendation widget , more than 100 of them earn on personalized native Relap.io ads . We know how users behave and what texts they read. This helps us create user segments of interest. We can recognize the user on any site and show him an advertisement, regardless of what he is viewing - a specialized website about tourism or news.

We divide the money from advertising in half with the site. The better we select segments, that is, the more accurate data on user behavior on different resources we collect, the more money the native advertising will bring to the site and the more we will earn.

So knowledge of user behavior on different resources helps us increase media revenue.

Read the material on Habré , where our lead developer talks about how we analyze the texts of media platforms.