International Press Institute (IPI) presents one of the most exhaustive studies on the current capacity and professional needs of digital journalism in Turkey. Prepared by IPI’s Turkish National Committee Vice President Emre Kızılkaya and journalist Burak Ütücü, the 68-page report evaluates the digital impact of Turkey’s independent media outlets, based on millions of data points in various mediums and metrics.

See the full report below.

DOWNLOAD FULL REPORT

---

---

Executive Summary

As 2020 closed, dozens of journalists remain in Turkey’s jails, with hundreds are on trial with baseless indictments that demand heavy imprisonment and hefty fines. Thousands are unemployed or practice self-censorship. Authorities cut the ad revenues and issue millions of liras in financial penalties targeting newspapers and broadcasters over their critical reporting. Dozens of outlets have been shut down either by their intimidated owners or directly through government decrees. Meanwhile, the daily circulation of Turkey’s best-selling newspapers has dropped below 200,000, and television stations are shackled by the highly-politicized state regulator.

Even as thousands of news websites and articles remain blocked by authorities who increasingly seek to exert control over the internet, most recently with the passage of the new Social Media Law in October 2020, the digital domain remains the primary refuge of independent journalism. But how does independent media perform on the internet and how does it compare to the reach of pro-government media? Do the media have the tools and skills necessary to flourish? Is there a ready audience? Can it secure the necessary revenue streams? What barriers are there to independent media and how do the policies of the gatekeepers of the internet —Google, Facebook, YouTube, and Twitter— affect the viability and potential of the digital news producers?

This report explores these questions and presents a comprehensive overview of the digital impact, capacity, and needs of independent outlets. It aims to help media stakeholders who seek to promote independent quality journalism in Turkey, in accordance with IPI’s mission to defend media freedom and the free flow of news wherever it is threatened. Hence,

• we studied 28 national and local outlets in Turkey, representing the editorial, geographic, and sociodemographic diversity, to compare the digital footprint of independent publishers’ in various mediums and metrics with pro-government publishers.
• We analyzed their overall digital reach with a representative panel of 16,104 users by utilizing a comprehensive dataset on Turkey’s digital audience. The mixed-method that also uses industry-standard on-site tracking provided us with detailed information about the monitored news websites’ “real users”.
• We collected 84,430 rows of data on Google’s search engine and Google News in news-related, high-volume, and trending queries in November 2020 to construct the most extensive dataset on the visibility of leading Turkish-language news outlets on one of their primary digital channels governed by the algorithms of the U.S.-based tech giant.
• We acquired a dataset based on millions of interactions with Turkish news articles on Facebook, Twitter, and Instagram. We mapped the monitored news outlets’ social media engagement performance in January-November 2020 and we listed the news articles with the highest engagement. We also analyzed YouTube’s algorithmic handling of independent and pro-government news outlets’ videos, while also looking into Twitter’s echo chambers.
• We conducted in-depth interviews and surveys with 20 independent media executives for a qualitative assessment of their attitudes and needs to produce and distribute quality journalism on digital. We also talked with journalists who recently became individual publishers on various digital mediums, including the web, podcasts, and newsletters.

Based on this rich set of information, our main findings are:

• Independent media’s digital reach (33.5 million users) is catching up with the pro-government media’s (47.8 million users). While the latter group’s reach has stalled recently, independent outlets continued to expand their digital user base rapidly.
• Independent outlets present a more fragmented media landscape compared to the highly-centralized and concentrated pro-government media. They also still lag behind the pro-government media in specific demographics, including women and youth.
• Independent media outlets receive 16.5% more interactions on social media and are closer to breaking through the echo chamber They dominate the pro-government outlets in almost all dimensions from their follower growth to the number of viral content, and on all platforms. For instance, they have five times more engagement on Facebook, and they reach a more diverse audience of news consumers on Twitter, while pro-government outlets function in what can be described as a filter bubble.
• In high-volume, news-related Turkish keywords and trending queries of the day, 90.6% of Google’s Top Stories slots were given by the search engine’s algorithm to three pro-government media outlets. The distinct algorithm of Google News appeared slightly less biased as it highlighted the pro-government media 73.8% of the times against the independent media’s 26.2%. These figures demonstrate that Google provides Turkey’s independent media outlets significantly lower visibility, despite their broad digital reach, robust social media interactions, and faster growth compared to the pro-government outlets.
• Google search results also amplify pro-government outlets’ hate speech and disinformation targeting democratic civil society. The company’s algorithmic preferences in Turkey remain in stark contrast with its standards in other countries, such as the U.S. where users are presented with a significantly more diverse menu of news sources led by highly reliable outlets, as well as more local publishers.
• YouTube’s recommendation algorithm mostly keeps the viewers of Turkey’s pro-government media outlets in an echo chamber. However, the same algorithm frequently channels the independent media’s viewers to the pro-government media’s news videos. YouTube’s one-way “algorithmic homophily” affecting the Turkish news video ecosystem needs to be investigated more thoroughly.
• Independent media executives report that most of their visitors come from social platforms and search engines. They think that Google incentivizes clickbait and ignores the violations of the largest news outlets, while Facebook favors “tricksters and trolls.” Some of them also criticized Twitter for not enabling Turkish news publishers to monetize their content more efficiently. Turkish digital platform Dergilik, too, is a source of concern as a gatekeeper, particularly for local media outlets.
• Only half of independent media outlets expected a net profit in 2020. 85% of their staff are in editorial teams, as they lack human resources, especially in tech-related departments. Although 60% of national outlets’ incomes are solely based on advertising, some are on their way to diversifying their revenue sources with digital subscribers, as they keep experimenting with New Media formats. Meanwhile, a lack of competition is a profound problem in local media.

In the conclusion, the independent media‘s digital reach is already significant and provides enormous potential for the future. However it is being held back by a combination of a lack of digital skills and resources to exploit their digital potential, and, it would appear, by social media algorithms that promote and prioritize pro-government media at the expense of independent and diverse views.

In order to address these challenges we recommend:

1. Connect and unite digital journalism players: Turkey’s independent media landscape is too fragmented. Those who are committed to quality journalism should unite and combine their forces. Such a nationwide coalition of journalism would not only share digital newsmaking skills and resources with each other for more effectiveness and efficiency but can also engage with digital giants and the government as an alliance to demand fair treatment by digital platforms.
2. Provide key skills and resources for digital media: Providing the tools for sustainable journalism, in the long run, should be a priority for media development. Turkey’s digital media have great potential, but they need access to the human resources and services, such as digital product development and business operations, that can enhance their digital competitiveness. The new structures and parameters for effective resource-sharing and fair access to the services and products required for quality journalism should be determined through more dialogue and cooperation between all stakeholders.
3. Create an ecosystem for independent digital media to flourish: The negative impact of search engines and social networks on Turkey’s independent media must be monitored and more effectively countered. Platforms, mainly through their algorithms, must treat news outlets transparently and with fairness, prioritizing the public interest in their decisions that affect the distribution of news content. Transparency over algorithms and take-down requests are ever more important given the new demands made on the digital platforms by the October 2020 amendments to the law on social media.  Efforts must continue to improve the negative political environment for independent journalism in Turkey, which is characterized by self-censorship and politically driven prosecutions.
4. The European Union is currently developing its regulatory rules for digital platforms through the Digital Services Act. Principles and policies endorsed will impact regulatory regimes in candidate countries and beyond.  It is essential therefore that the EU strikes the correct balance in ensuring a fair, open, transparent, and viable platform for independent media to operate.

 

---

 

Foreword and Acknowledgements

I was the Hürriyet newspaper’s digital editor when I had an interesting chat with my counterpart at CNN Türk television on a sunny day in June 2016. Both of us were puzzled because Reuters Institute’s Digital News Report, which had just been published, ranked CNN Türk as the most popular online news outlet in Turkey, while Hürriyet was listed only on third place.

The two news brands were part of the Doğan Media Group then, which was among the largest media corporations in Europe, reaching millions of readers, viewers, and listeners each day. We were puzzled because both of us, as the managing editors of these two outlets, knew that Hürriyet’s website reached an average of 30 million unique users each month, while CNN Türk could only get around 10 million. The difference for pageviews was even more significant as CNN Türk’s volume was only one-tenth of Hürriyet’s in this metric. After all, Hürriyet had even more pageviews than the New York Times at the time.¹

We were not the only ones who were puzzled, though. From Turkey’s leading publishers to its biggest advertisers, all relevant parties in the digital news ecosystem knew well that Hürriyet’s digital footprint was many times bigger than CNN Türk’s. After all, these parties had access to real-time traffic data through the dashboard of an independent auditor hired by the publishers’ and advertisers’ umbrella organization, the Interactive Advertising Bureau (IAB) in Turkey. Simply put, the YouGov survey that the Reuters Institute’s report used was flawed probably due to polling errors², but everyone quickly forgot about it as only a month after that sunny day, a military coup attempt happened.

Even though the coup-plotters raided our offices as we still insisted to do pro-democracy journalism³ in July 2016, the government used the event as an excuse to maximize pressure on independent outlets like ours under emergency law, which ultimately forced Doğan to sell all its media assets to a staunchly pro-government corporation, Demirören, in 2018. As I wrote in the Nieman Reports⁴, I resigned from Hürriyet months after the sale and became the project editor of Journo.com.tr, a non-profit digital platform for next-generation journalists, supported by Turkey’s Journalists’ Union (TGS) and the European Union.

In a webinar organized by TGS’ Media Academy on 11 December 2020, Kadri Gürsel, the IPI’s Executive Board Member and the president of its National Committee in Turkey, described the sale of Hürriyet and CNN Türk as the final step in a decade-long government campaign to control the mainstream media. President Recep Tayyip Erdoğan’s administration had issued Doğan, Turkey’s largest media group, a $2.5-billion tax penalty in 2009, months after Hürriyet published a corruption story that embarrassed his ruling Justice and Development Party (AKP). By arresting, intimidating, and targeting journalists and publishers over the following years, Erdoğan achieved almost total control over Turkey’s largest media outlets by 2018. His allies who have business interests in other sectors inherited the digital assets of their newly-acquired newspapers and televisions, too, expanding pro-government propaganda and disinformation into New Media.

As Gürsel stressed, Hürriyet could be seen as Turkey’s last “mainstream” media outlet because it provided a large platform for various views and perspectives. The mainstream media outlets of the past, including Milliyet where Gürsel worked for years, showed this quality not only in their editorial policy, such as having opinion columnists representing different political or social groups but also in their newsroom diversity, which enabled them to reach a broader audience.

“In the largest newsrooms of Turkey in the past, editors with -let’s say- social democratic views could sit together with nationalists or conservatives around the same table, having a continuous conversation on how to cover news stories. Despite their differences in ideology or elsewhere, they were applying journalistic standards in the end when it comes to editorial decisions,” Gürsel said.

Despite all their flaws and shortcomings, the larger and centralized media outlets in Turkey, which were formerly known as the “mainstream,” were able to raise great journalists like the former IPI board member and Milliyet editor-in-chief Abdi İpekçi. Gürsel remembered İpekçi for his principled stance against the “harmful currents” in politics, society, and the economy that undermined independent journalism in Turkey for years to come.

From the age of Ipekci, Turkey’s mainstream media had been the primary home to world-class professionalism, editorial diversity, and assets to produce high-quality journalism, including human resources and financial means. The AKP’s years-long campaign, however, led to the slow-motion destruction of this home over the past two decades.

Today, journalism still survives in Turkey in a less centralized, much more fragmented, and almost wholly digital media landscape. As Gürsel pointed out, it may be wrong to describe today’s independent outlets, who try to cling to journalism despite all the challenges, as the “alternative” media. “More accurately, they can be labeled as the kernel of a new ‘mainstream’ media,” he added.

We cannot understand where Turkey’s independent media stands now and is heading toward by using patchy and sometimes misleading data on our peculiar news ecosystem, which is the case at the moment. This report aims to be a comprehensive, qualitative and quantitative assessment of this new media’s digital capacity, impact, and needs.

As the evolution of these outlets toward or away from becoming “the new mainstream” will fundamentally affect the future of quality journalism in Turkey, we believe that the data-driven insights and recommendations in this report can be useful for all stakeholders who seek to know more about independent news outlets and how to support them to become sustainable and to fulfill their democratic role.

The report could not be possible without the European Commission’s financial support and the indispensable help from the IPI Vienna staff to whom we are grateful. For their valuable contribution, I would also like to thank all journalists and media executives whom we interviewed, as well as to my fellow researcher Burak Ütücü, data engineer Emrah Yılmaz, Gemius’ Country Manager İdil Kesten, and Somera’s CEO Gencay K. Evirgen.

Emre Kızılkaya

Vice President of IPI’s National Committee in Turkey

---

Methodology

Our research has a multi-method approach to study qualitative and quantitative aspects of the digital capacity, impact and needs of a set of Turkey’s independent news outlets in relation to centralized media corporations that were formerly known as the “mainstream”.

Whom We Studied

Group A

There are dozens of independent media organizations in Turkey at the moment. To get a sample as representative as possible in the scope of this study, we selected 14 national outlets, which are among the leaders in various mediums and metrics. These national news outlets and the thematic groups that they represent, categorized as “Group A” of our sample, include the following:

• Sözcü: The leading omnichannel independent news platform with an established brand in print journalism. The reach and size of its website and digital operations are comparable to the outlets of the biggest (pro-government) media corporations.
• Cumhuriyet, BirGün, Evrensel, Yeniçağ, Karar: Established newspapers with an increasing focus on digital transformation. Their primary audiences represent different political, social, and cultural groups, including secularists (Cumhuriyet), social democrats (BirGün), nationalists (Yeniçağ), and conservatives (Karar). Their editorial focus is also thematically diverse. For instance, Evrensel’s coverage is more focused on labor and human rights.
• OdaTV, Gazete Duvar, T24, Diken: Digital-native news outlets whose primary focus is web journalism, mostly conducted through techniques and formats of classical newsmaking built around text. Many of their employees are journalists with a background in newspapers. Most of these outlets have also been experimenting in new formats and mediums in recent years, including video and podcasts.
• Halk TV and Tele1: The leading news televisions with critical reporting and an increasing focus on digital transformation. They recently turned their websites into news portals and invested more in digital video by focusing on off-site platforms including YouTube.
• Medyascope: A digital-native news and opinion platform focused on video and social media. Many of their employers are journalists who were previously employed in television news. It received IPI’s Free Media Pioneer Award in 2016.
• Teyit: An independent fact-checking platform, operating as a social enterprise. Producing content in multiple digital formats, it is the official fact-checking partner of Facebook and TikTok in Turkey.

All these independent media outlets, except Teyit, are based in Istanbul, which is the traditional center of today’s media. Teyit is based in the capital city of Ankara.

Group B

Although the focus of this study is independent media outlets, we used seven other leading news organizations as a benchmark. Categorized as Group B in our study, they include some of the largest outlets of the old “mainstream” and have links to the ruling party or under its political influence in varying degrees⁵:

• Hürriyet was once the most influential daily newspaper. It operated the most popular news website in Turkey as of 2018, when the staunchly pro-government conglomerate, the Demirören Group, bought it and its sister outlets in print, television and radio. Since then, Hürriyet’s critical reporting has been largely muted.⁶
• Sabah is considered as the Erdoğan administration’s flagship newspaper as part of the Turkuvaz Medya Group, which also owns other newspapers, televisions, and radios. The “official” owner of the group is currently Zirve Holding, which is fully-owned by Ömer Faruk Kalyoncu, a businessman whose companies were awarded billions of liras-worth of public tenders in construction, energy and infrastructure sectors⁷. However, the effective control of Sabah newspaper’s editorial policy lies in Turkuvaz board member Serhat Albayrak, the brother of Erdoğan’s son-in-law Berat Albayrak.
• Yeni Şafak and Akşam newspapers are controlled by two families, the Albayraks⁸[who are not related to Berat Albayrak’s family] and Yesildags⁹, whose members have active roles in the AKP, as well as vested interests in various business sectors. Both of them are from the Black Sea region, which also houses Erdoğan’s hometown.
• Haber7 is a news website owned by Beyaz Holding, which also owns Kanal 7 and Ülke TV televisions, as well as Radyo 7. The corporation’s board members were put on trial and its founder was convicted in the Deniz Feneri E.V. corruption case in Germany in 2008. Since then, Beyaz Holding was reportedly awarded hundreds of millions in Turkish liras through public tenders by AKP-controlled municipalities.¹⁰
• Habertürk’s newspaper was shut down in 2018, but its television and website remain among the top news outlets in their categories. Although it is still home to a limited number of critical journalists and refrains from an editorial policy as partisan as the outlets listed earlier. The most recent example of this stance was the state-controlled broadcasting watchdog The Radio and Television Supreme Council’s (RTÜK) decision to heavily sanction Habertürk for hosting a main opposition party MP who harshly criticized government policies live on air. ¹¹, it still has a remarkable pro-government bias in its coverage of key events. Its owner is the magnate Turgay Ciner’s holding whose main investments are in mining, energy, real estate, and tourism.¹² Habertürk’s news website consistently ranks near the top in the list of the most popular news websites in Turkey.
• NTV, like Habertürk, has its roots in broadcast news and tries to display a relatively more balanced editorial stance compared to heavily-biased outlets like Sabah. It operates as part of the Doğuş Media Group, owned by the Şahenk family, who also has business interests in automotive, construction, tourism, real estate, energy, entertainment, and retail sectors. NTV’s news website consistently ranks near the top in the list of the most popular news websites in Turkey.

Group C

We also looked at seven local news organizations from different parts of Turkey to cover all the regions of the country, and categorized them under Group C:

• Pusula from the Western Black Sea province of Zonguldak
• Haber61 from the Eastern Black Sea province of Trabzon
• Manavgat Son Haber from the Mediterranean province of Antalya
• Pusula from the Eastern Anatolian province of Erzurum
• Tigris Haber from the Southeastern Anatolian province of Diyarbakır
• Yayla Haber from the Central Anatolian province of Çorum
• Dokuz Eylül from the Aegean province of Izmir

Pusula (Zonguldak), Haber61, Yayla Haber and Manavgat Son Haber were the four local news organizations from Turkey selected by Google’s Journalism Emergency Relief Fund (JERF) in July 2020.¹³ We picked the remaining outlets because they were among the few established news brands in their local communities with a more significant digital reach and impact as observed in our pilot study.

What We Studied

The first part of the study is a quantitative look into Turkey’s top digital media outlets’ current digital reach.

After providing up-to-date information about the audience size and sociodemographic reach of the media outlets tracked in three groups, we delved deeper into their digital impact by collecting and analyzing data about the off-site engagement and visibility of their news content. Our focus was their activities on social networks and Google-owned platforms, which are the primary sources of traffic and digital revenue for almost all outlets in our sample.

The second part of the study has a qualitative approach to understanding what was revealed in the first part, explaining the “whys.”

By interviewing journalists and managers of independent media outlets in our sample, we tried to depict various aspects of their organization, resources, limitations, plans, strengths, weaknesses, tactics, strategies, and needs that affect their digital impact and potential.

The last part discusses all the findings, provides analysis and actionable recommendations for the stakeholders, including donors and platforms, to support independent journalism in Turkey on the way to digital sustainability of the news ecosystem.

How We Studied

The first part of the study has three dimensions: (1) Overall digital reach of Turkey’s leading news outlets; (2) their impact on the predominant search engine, Google, and (3) their impact on social platforms. We used industry standards and scholarly methods to collect and analyze the data in all three dimensions:

Overall Digital Reach

Several companies, including the U.S.-based Alexa Internet and SimilarWeb, claim to provide accurate statistics for online traffic and digital engagement globally. However, our own experience in the news business and the media executives we talked to suggest that –while such services tend to reflect the Turkish digital media trends relatively better– most of the net figures that their algorithms and panel analyses provide are highly inaccurate, as they are not designed for or focused on the Turkish media landscape.

As self-announced data from the media companies are also not wholly reliable, we cooperated with the Warsaw-based international research and technology company Gemius, which was used for years by IAB as Turkey’s leading publishers’ and advertisers’ joint platform with agreed-upon standards and methods to measure the media outlets’ digital traffic in the country.

Gemius, a member of IAB Europe, has the ISO 27:001:13 quality certificate and follows the European Society for Opinion and Marketing Research (ESOMAR) guidelines. With the “Gemius Audience” digital media study, the institution regularly uses a representative panel of 16,104 people (PC/mobile) throughout Turkey to measure the digital media consumption and behavior of online users. Gemius combines the user-centric data with its site-centric measurements based on the tracking codes working on the participating media outlets’ websites. Its “real user” metric, created by an algorithm that operates with the mixed-method data, singularizes a unique consumer who uses multiple devices to access digital news.¹⁴

For this research, we acquired the Gemius Audience dataset for October 2020 to publicly reveal for the first time the overall digital reach and real users of each news outlet in the three groups that we studied. The set includes data on sociodemographic audience composition while providing our three groups’ reach specifically to various segments in Turkey based on age, gender, and socioeconomic status (SES).

Search Engine and Social Platforms

According to the Trust Project’s Sally Lehrman whom the Reuters Institute interviewed, many news executives in the U.S. felt like they were held “captive to social media and search engines and just the digital environment in general”.¹⁵ This is also the dominant view among news executives in Turkey, whether they work for independent or pro-government media, because the majority of their outlets’ digital traffic originates from Google and social platforms, as our interviews for this study also confirmed.

To measure the visibility of Turkish news outlets on Google¹⁶, we used a custom Python code to collect three sets of news-related queries every hour from November 6 to December 6, 2020. As in a recent academic study in this field¹⁷, the virtual machines in our cloud-based research infrastructure acted as bots and trackers to collect the data by emulating a newly-installed Chrome browser on default settings and “Incognito” mode on a PC and a mobile phone to avoid personalization of search results.

Firstly, we collected data on Google’s core product, its search engine. On Google Search, we regularly queried two of the largest-volume news-related keywords, which are “son dakika” (breaking news) and “haberler” (news). These two keywords, driving more than 1.6 million visits to Turkish news websites every day, are more closely associated with the most avid news readers.¹⁸

In order to understand what casual readers were “incidentally exposed”¹⁹ to by Google-curated news stories, we also looked into the trending queries. To do this, we created another virtual machine to fetch the regularly updated public data on Google Trends and scraped the search engine result pages (SERPs) for these specific queries to see which news outlets are favored by Google on trending searches. Like in “son dakika” and “haberler” queries, we recorded the SERP positions (in and out of the Top News carousel), row orders, titles, URLs, and other attributes of the results for the trending search keywords.

On Google News, without using the search function, we scraped the homepage every hour for a month, recording the titles, row orders and URLs for five news articles in each of the top five “Headlines” boxes on this platform.

As a result, we constructed the structured dataset of 84,430 entries that formed the basis of our analysis by capturing relevant data on the visibility of Turkish news outlets on Google’s search engine and Google News in November-December 2020.

For social media, on the other hand, we cooperated with Somera, an Istanbul-based technology developer who has industry expertise on social media monitoring. The company shared with us a database, built by its own custom tool, of various engagement metrics for each media outlet that we probed for every week from January to November 2020. The number of followers, shares, comments, replies, video views were among the primary categories of interactions regularly measured on Twitter, Facebook and Instagram throughout this period. We also listed the Turkish news content that received the greatest amount of social media engagement in the same period.

We conducted a separate analysis on Twitter to understand the degree that independent and pro-government outlets serve on this social network as echo chambers or not. We collected 112,482 Twitter connections around both types of outlets’ content in January by using NodeXL Pro, a licensed Microsoft Excel add-in for social network and content analysis, developed by the U.S.-based Social Media Research Foundation. We visualized the data with Gephi, using the ForceAtlas2 algorithm.

For YouTube, we manually collected some of the data, like the number of followers of the media outlets’ channels. We also collected the data on the algorithmic recommendations of YouTube with NodeXL Pro. We limited the search to two levels with three videos per level, mapping a total of 69 videos from the studied outlets’ three most popular videos on YouTube. We graphed it with Harel-Cohen Fast Multiscale, but any other algorithm would produce the same results because the sample size is small and other parameters are not complex.

Surveys and Interviews

We conducted semi-structured interviews with the digital executives and managing editors of independent media outlets in November-December 2020.

The participants from the Group A national media outlets were Medyascope editor-in-chief Ruşen Çakır, Evrensel editor-in-chief Fatih Polat, Teyit founder Mehmet Atakan Foça, Sözcü chief digital officer Erhan Acar, Diken founder Harun Simavi, BirGün Internet Director Hakan Demir, Oda TV managing editor Barış Terkoğlu, Cumhuriyet managing editor Mustafa Büyüksipahi, Karar managing editor Fuat Atik, T24 managing editor Sertuğ Çiçek, Tele 1 Editors Board Member Oğuz Güven, Halk TV digital editor Fırat Yeşilçınar, and Gazete Duvar director of marketing and advertising Fırat Özdemir.²⁰ The Yeniçağ newspaper declined to comment.

The participants from the Group C were all editors from local media outlets, namely Murat Attila from Dokuz Eylül, Ali Yılmaz from Tigris Haber, Cihet İncesu from Pusula (Erzurum), Adem Polat from Manavgat Haber, Doğan Alayoğlu from Haber61, Hacı Odabaş from Yayla Haber and Bayram Tomakin from Pusula (Zonguldak).

Before the in-depth interviews, the participants filled in a survey with 36 questions on various aspects of their digital news operations including 13 questions that they rated on a 5-point Likert scale. All surveys and interviews, which took around 45 minutes on average, were conducted online due to the safety measures amid the COVID-19 pandemic. The survey included questions on the related media outlets’ organization, human resources, finance, technology, digital performance, perceived needs, and their relations with, as well as attitudes about platforms.

We used Datawrapper and Flourish for visualization and TwoTone for the sonification of our data. As we wrote this report, we also started to develop a public dashboard that everyone can use to monitor various real-time data affecting the Turkish news media landscape and the main platforms that they distribute their content.

 

---

Findings

Overall Digital Reach

Turkey’s online population stood at 58,546,876 real users in October 2020, according to the Gemius data. Alongside Google, Facebook, YouTube and e-governance services, the Top 20 list of the most visited websites in Turkey regularly features an average of five news organizations.

Among the Group A outlets that we defined, Sözcü has the largest monthly reach with an audience penetration of 38.67% as of October 2020²¹, comfortably remaining one of Turkey’s most popular websites:

There is a key factor to note here: Firstly, some of these outlets are not focused on their news websites. For instance, Medyascope’s priority is not to draw visitors to its website, but to engage with them on offsite video platforms, particularly YouTube and Periscope. Something similar can be said for others. TV channels like Halk TV and Tele 1 are more focused on YouTube, while the fact-checking organization Teyit’s success lies not in merely attracting traffic to its website, but in dispelling misinformation and disinformation on its primary source  – social media platforms, such as Facebook. This was one of the reasons that we also gathered data on independent media’s reach on these social channels in the latter part of the report.

The Gemius dataset shows that two of the Group B outlets, Hürriyet and Sabah, reached more than half of the Turkish digital audience in October 2020²²:

Comparing the two groups’ total reach, Group A outlets penetrate 57.3% (33.5 million users) of Turkey’s online audience, while Group B outlets reached %81.8 (47.8 million users).

We did not conduct a more comprehensive affinity analysis between and in these two groups. However, based on the characteristics of the reader behaviors among the supporters and critics of the ruling AKP, it can be argued that independent outlets’ reach is at least as diverse as – if not even more than – Group B outlets, although the former group’s current audience penetration seems lower than the latter. This reasoning is based on the simple fact that the Group B outlets’ editorial policies are almost monolithic – inspired by a right-wing or center-right, pro-state ideology – which limits the perspective of its news coverage to conservative angles, while Group A includes outlets that are more varying in this sense, having secularist/Kemalist, moderate Islamist, Turkish nationalist, socialist, Kurdish rights-focused and politically-neutral views.

In this sense, it is remarkable that the total reach of two independent newspapers that address a similar audience base, Sözcü and Yeniçağ, has already reached the level of the largest pro-government newspaper, Hürriyet, as of October 2020. Similarly, if the left-wing independent media outlets, such as Cumhuriyet, BirGün, Oda TV, Gazete Duvar and Evrensel, are considered as a whole, it is seen that this group’s combined digital reach is greater than each and every Group B media outlet except the top three – Hürriyet, Sabah, and Habertürk.

The gap between Group A and Group B outlets’ digital reach varies among sociodemographic groups in Turkey²³. The median of this variation is 24.8%. In this regard, independent outlets outperform the pro-government media in the 45-54 (19.2%) and 55+ age groups (20.3%) and among males (21.6%). However, their most significant underperformance versus Group B outlets is in the 18-24 (28.7%) and 15-17 (27.4%) age groups, and among females (27.7%). The two groups did not divert from the median significantly in any of the SES segments.

Among local media outlets in the Group C, Haber61 website in the Black Sea province of Trabzon is the only news organization with a wide-enough digital reach that is above the threshold of the monitoring radar. According to Gemius, this website was visited by 587,600 real users in October 2020, reaching roughly 1% of the online audience in Turkey.

Social Platforms

The graph below, which compiles all Facebook, Instagram and Twitter interactions of the Group A outlets excluding Sözcü newspaper, shows that the first and the second wave of the COVID-19 pandemic heavily affected social media engagement of Turkish news media.²⁴

For most weeks, Cumhuriyet daily was the leader of social media engagement among the 13 outlets. T24, Tele 1, BirGün, Halk TV, Oda TV and Teyit are among the outlets that consistently outperform on social media in comparison with the digital reach of their websites. Yeniçağ, on the other hand, is a negative outlier on this count, as its social media engagement remarkably falls behind its website’s reach.

We graphed Sözcü’s social media interactions separately, as it has an enormous volume that overshadows other Group A outlets, as well as most of Group B. Considering the dramatic drop in late summer, it can be presumed that a significant proportion of Sözcü’s social media interactions are sourced in the opinion articles of its popular columnist Yılmaz Özdil.

Yeni Şafak has been the leader among Group B outlets in social media engagement throughout the year, excluding three weeks that NTV passed it and one week that Haber 7 did so. Still, none of the pro-government Group B outlets can reach the level of social media engagement of Group A’s leading outlet Sözcü.

When looking at the graph above, it should be noted that Instagram video views with — a weekly average of 2 million — had an outsize influence on Yeni Şafak’s total social media interactions. None of the other Group B or Group A outlets had such a dramatic outsizing factor on any platform or metrics. Even with this distorting effect, the Group A outlets have had 16.5% more interactions on social media than the Group B outlets:

Group C, which includes local media outlets across Turkey, had several times lower volume of social media interactions compared to the national outlets in Group A and Group B, as could have been expected. Because of Facebook API’s limitations, we first looked into their Twitter interactions from January to November 2020. Only three of the Group C outlets — Tigris Haber, Haber 61 and Dokuz Eylül — received more than 100 weekly interactions on Twitter throughout this period:

One of the reasons for this limitation is that the total data volume is too big to fetch externally, as all links of the media outlets — whether they were shared by the media outlet itself or by any social media user in Turkey — had to be analyzed. By using Facebook’s own tool CrowdTangle, we overcame most of the obstacles and analyzed each outlet’s performance on this platform from 1 January – 15 December 2020.

The Group C, which includes local media outlets across Turkey, had several times lower volume of social media interactions compared to the national outlets in Group A and Group B, as could have been expected. Because of Facebook API’s limitations, we first looked into their Twitter interactions from January to November 2020. Only three of the Group C outlets — Tigris Haber, Haber 61 and Dokuz Eylül — received more than 100 weekly interactions on Twitter throughout this period:

One of the reasons for this limitation is that the total data volume is too big to fetch externally, as all links of the media outlets — whether they were shared by the media outlet itself or by any social media user in Turkey — had to be analyzed. By using Facebook’s own tool CrowdTangle, we overcame most of the obstacles, and analyzed each outlet’s performance on this platform from 1 January – 15 December 2020.

As seen in the table, the independent outlets’ in Group A have five times more Facebook interaction than Group B’s pro-government outlets. Group A’s pages grew almost 10 percent, while Group A’s page likes shrank 0.61% in the past year. Local media outlets stand between these two groups both in weekly interaction rates and the page growth metrics.

All Group A outlets, except Oda TV and Karar, grew on Facebook in 2020. The daily Sözcü had the greatest number of new Facebook followers (81,800), while the annual growth rate was highest at the video-oriented Medyascope (45%).

Among the Group B outlets, only NTV’s and Haber 7’s Facebook pages grew in 2020 (2% for each). The greatest fall was at Hürriyet, which lost more than 85,000 followers in less than one year, shrinking 2.7%.

The local media outlets of Group C also boost their followers by 15%. Haber 61, which has the largest Facebook page in this group, grew 10%, while Dokuz Eylül’s interactions raced with this outlet for several weeks.

Although the focus of this report is not the particular news stories with the widest reach, looking into the most popular content specifically shared by these media outlets enabled us to get deeper insights. Overall, there was 61 content from the three groups of outlets that attracted at least 10,000 interactions each on Facebook, Twitter, and Instagram from January to November 2020.

The most popular post, which received 119,053 interactions, was a tweet by the Group A outlet Cumhuriyet daily. It was reporting Turkish actor Şevket Çoruh’s response to pro-government Sabah’s columnist Mevlüt Tezel’s recent statement that criticized private theater owners of “not saving money before the pandemic to use during the crisis.” As a theater owner himself, Çoruh criticized the ruling politicians in response, sarcastically asking why the Turkish government had to call its citizens to donate money to be used for the pandemic.

In the Top 10 list of most popular social media posts of news outlets, the Group A outlet Sözcü daily has six entries, four of them being journalistic content critical of the government policies. A post by Sözcü’s popular journalist İsmail Saymaz is also in the list, sharing a critical news story about the pandemic-era government policies (Only Cumhuriyet’s and Saymaz’s posts were from Twitter, while the rest of the Top 10 were shared on Facebook).

The remaining content is the only one from a Group B outlet that managed to enter the Top 10 list. It is a non-political news video created by Demirören News Agency and shared on daily Hürriyet’s Facebook page, telling about an “engineering innovation” of a retired car mechanic in Istanbul.

Beyond the Top 10, the full list of 61 viral content provides more insights. Publishers of most viral content (50 articles) are seven independent outlets. Five of them are national outlets — representing various social and political leanings from the right to the left — namely Sözcü, Cumhuriyet, T24, Duvar, and Teyit.

The presence of the latter outlet in this list for social media prominence is especially important, as Teyit is Facebook’s official partner for fact-checking. The viral stories that were fact-checked by Teyit this year included issues such as the legalization of homosexual relationships during the Ottoman era to the Wayfair conspiracy theory in the U.S.

Except for the two Sabah articles that targeted the main opposition party and its leader, most of the viral content from the Group B outlets were non-political human interest stories and articles about the coronavirus pandemic.

It may be encouraging for free press advocates to see two local outlets in Turkey managing to enter this viral content list. However, it should be noted that both viral Facebook content of Haber 61 in the list were merely statements of support to the Black Sea province’s popular football team, Trabzonspor. The only other content, produced by Manavgat TV from the Mediterranean province of Antalya, was a heartwarming video news story about local security forces who helped an old farmer to harvest his field during the pandemic.

We also conducted a network analysis on Twitter for Group A and Group B outlets in January 2021. The following graph includes more than 30,019 nodes, which shows the media outlets and the particular users who interacted with them through retweets, mentions, follows, etc. in this period. Such interactions created 112,482 connections.²⁵

These results suggest that Turkey’s independent outlets reach a more diverse audience of news consumers on Twitter, while pro-government outlets function in what can be described as an echo chamber. Still, even a follower of a staunchly pro-government media outlet like Haber 7, which is on the edge of the echo chamber, can reach an independent outlet by a few “jumps” through common connections.

Group A’s Sözcü, Oda TV, Cumhuriyet and Yeniçağ, and Group B’s Habertürk and Hürriyet are closer to perform a bridging role between these groups in this regard. Teyit, Medyascope, Evrensel and Karar, on the other hand, have an important function as “outer connectors” of diverse audiences within the independent media audience.

As can be seen, there is a certain degree of polarization on Turkish Twitter regarding news consumption, but it is not decisive, as a significant portion of the audience has access to a diverse set of sources through these “inner bridges” and “outer connectors.” Independent media outlets are closer to the audience’s center of gravity.

These findings can also be considered in the context of the Turkish government’s latest move to block Twitter for failing to appoint local representatives as required by the new Social Media Law.

When we look into the Twitter interactions of the selected local media outlets in Group C in the same period, we see that — beside Haber 61 again — the content of Tigris Haber from Turkey’s southeast and Dokuz Eylül from the west performed better. Each of the other outlets in the group had less than 100 weekly interactions on Twitter.

We also studied the Turkish media outlets’ presence and interaction on YouTube, which is a key platform for many of them, including the broadcasters like Halk TV and next-generation video-based news outlets like Medyascope.

In the following graph, the size of a circle is proportional to the number of subscribers of the related media outlet’s YouTube channel as of November 2020. Blue is used for Group A outlets and red for Group B.

Total video views or the number of subscribers of a YouTube channel can be misleading in determining interactions by themselves. Hence we created two metrics to get a more accurate insight: The first is the “engagement” score, which we found by dividing a channel’s total video views by the number of published videos.

The second metric is the “high-quality reach,” which we found by dividing the number of a channel’s subscribers by 1/1000 of its engagement score. We picked this specific multiplier to create a high-quality reach score closer to the engagement score in order to put them on the same visualization as seen below:

Habertürk, Halk TV, NTV, Yeni Şafak, Medyascope and Tele 1 all have more than 200,000 YouTube subscribers. The first four outlets in this list have got the most (total) video views on YouTube.

If the broadcasters, which have a significant competitive advantage, are set aside, it is notable that the staunchly pro-government Yeni Şafak daily and the next-generation independent media outlet Medyascope received the greatest number of video views.

One of the reasons behind Medyascope’s success may be based on the fact that it is the most prolific creator of video news and commentary as per its daily production on YouTube. Cumhuriyet and Tele 1 follow Medyascope in daily video production.

While Evrensel, NTV, and the Group C outlet Zonguldak Pusula have the highest engagement per subscriber, Oda TV, Yeni Şafak and Teyit have the highest engagement per video.

The size of the subscriber base provides a media outlet with a multiplier effect for its user engagement. As a result, it can be seen in the graph that Habertürk, Halk TV, and Oda TV have the highest quality reach on YouTube, hence the greatest effective capacity. Sözcü’s latest initiative to establish a broadcast TV channel may boost its capacity on YouTube as well.

However, Oda TV’s presence in this list — the absence of NTV and Tele 1 — suggests that while having a broadcast TV may bring a significant advantage to a news media outlet on YouTube, it is not a sufficient condition. Neither having an “old” YouTube channel is required, as Oda TV reached more than 100,000 subscribers in one year. Doubtlessly, it is more about content than the outlet, when it comes to YouTube.

The Google-owned platform’s “rules of engagement” heavily affect the performance of a particular video or the channel of a media outlet. The paradigms are yet to be studied extensively for Turkish news content on YouTube and a thorough assessment including trust metrics and political discourse exceeds the limits of this research. Still, we wanted to briefly look into the issue of “algorithmic homophily,” which is described as a curating algorithm’s tendency to reinforce echo chambers by recommending the same or similar content to the users who share common characteristics.²⁶

We conducted this analysis by focusing on the top three YouTube performers in our three groups of media outlets. For each outlet, we listed their three most popular videos produced in 2020.

The first YouTube channel we studied was Medyascope’s, which scored a rare success in video reporting and news commentary as it managed to surpass many legacy publishers in a few years as a New Media startup.

Most of Medyascope’s top performing YouTube videos of the year feature the street interview format, giving the microphone to the common person and asking them for comments on political issues of the day, such as local elections. The top video, a street interview on the victory of the main opposition candidate in Istanbul metropolitan municipality elections, was watched more than 850,000 times.

When we graph the YouTube recommendations for each of the top three videos of Medyascope, a tree is shaped that most branches lead to Habertürk, a Group B outlet (green dots).

In this tree, Halk TV and Tele 1 are the only Group A (independent) outlets whose videos are recommended by YouTube to Medyascope’s viewers. Most of the rest are recommendations to small channels (yellow) focused on street interviews, one of them being Ahsen TV (brown), a staunchly Islamist/conservative publisher. BBC Turkish and DW Turkish (purple) have two branches.

The pro-government Yeni Şafak’s YouTube recommendations tree for its top three performing videos in 2020 is in stark contrast with the independent outlet Medyascope’s.

The most striking difference in the tree is that it is much more homophilous. Most recommendations are addressed to other Yeni Şafak content. The rest goes to other, smaller pro-government publishers (brown) or neutral publishers (yellow).

The recommendation tree on three levels has 31 videos, which were watched by more than 36 million people and got 170,000 likes. Reviewing the content, we see that Yeni Şafak’s three most popular videos feature statements by President Erdoğan and the late conservative politician Muhsin Yazıcıoğlu, as well as a human interest story, focused on an old father who takes care of his disabled son.

Most of the recommended videos lead the viewer to content with politically similar (from other Islamist politicians like Erdogan’s former leader, Necmettin Erbakan) statements or similar human interest videos (like the story of other older or disabled people who “thank God despite all the challenges in their lives”).

The 19 videos on Oda TV’s recommendation tree, on the other hand, were watched more than 48 million times and received 330,000 likes. This independent outlet’s YouTube channel includes original news footage and profiles. Their three most popular videos of the year include two interviews and the archive footage of court testimony from the Ergenekon trial.

The YouTube recommendations for independent media outlet Oda TV’s three most popular videos of 2020 were more similar to Medyascope’s than to Yeni Şafak’s in the observed period. Again, they usually lead viewers to the content of Group B outlets, including the staunchly pro-government A Haber. Purple dots represent entertainment content, including TV dramas.

Although more extensive research with larger samples is needed for a thorough analysis, the examples from the most popular videos of these three outlets suggest that YouTube’s algorithmic homophily works unequally in Turkey: While the viewers of independent media outlets’ videos are presented by YouTube with a set of more diverse recommendations politically and socially, the viewers of pro-government outlets’ are mostly left in their own echo chamber. Besides its potential harms in informing the public, this mechanism apparently boosts the performance of pro-government outlets by channeling more traffic -and ad revenue- to their content, while undermining the independent media.

Search Engine

As noted earlier, Google is the dominant search engine in Turkey with 84% market share. The surveys and interviews we conducted for this research confirm that the source of roughly half of the studied media outlets’ digital traffic was Google as of November 2020.

Millions of Turks get their news “incidentally” on Google when they search for information and are presented by the algorithm results that are usually dominated by news publishers’ content. As a result, Google’s gatekeeping algorithm — which regulates its organic search results, carousels and the Google News application’s curation — plays a key role in the visibility and discoverability of news content produced by Turkish media outlets.

As detailed in the methodology section, two specific news-related Turkish queries (“son dakika” and “haberler”) regularly channel millions of users to these outlets’ content everyday. Trending searches, on the other hand, are another substantial traffic source, although they change from one day to another.

While the first type of searches mostly brings the audience segments that can be called “news lovers” and “daily briefers,”²⁷ the second type primarily attracts “casual users” who end up in a news outlets’ website incidentally. Still, both types of users generate ad revenue for the visited website, so —as in YouTube — Google’s algorithmic choices for its search engine are important not only for the health of an online user’s information diet but also for the financing of a news publisher. Google’s news carousels and search results are also a marketing channel for publishers who are focused on reader revenue.

Emre Kızılkaya from IPI Turkey had reported in July 2020 that Google’s Top News carousel frequently features hate speech, propaganda and disinformation from pro-Erdoğan media outlets.  As seen in the example above, the problem persists as jailed businessman and prominent human rights activist Osman Kavala was targeted in yet another smear campaign by Sabah newspaper on December 18, and Google amplified the hate speech and disinformation once again by providing this outlet the two top slots of its Top Stories carousel. Kavala has been under arrest for more than three years with no evidence but the conspiracy theories spread by pro-government outlets.
Meanwhile, the screenshot below shows that this is true not only for local targets, like Kavala, but also international scapegoats. The Google searches on January 27, 2021, for George Soros led the Turkish users to conspiracy theory-laden pro-government news websites. The top result was an opinion article from Sabah, which warned the readers to “get ready for Soros’ new attempts” in Turkey by claiming that the businessman was behind the anti-government protests in Russia. The second story picked by Google was a news article by Hurriyet, which claimed on its title that “Bill Gates, Soros, and Rockefeller were accused of creating the virus that caused the COVID-19 pandemic.” No independent media outlet appeared in the top results on Google for this query.

To get an international benchmark for Google’s algorithm in picking the most highlighted media outlet, we looked into the top results in trending searches in the United States from October 25 to November 24, 2020. According to the data from Google Trends’ daily newsletter²⁸, a total of 225 U.S news outlets appeared as the top result in 541 trending searches in this period. Only eight of them were the top result for more than 10 times each: ESPN, CNN, New York Times (NYT), USA Today, National Public Radio (NPR), Deadline, CBS Sports, and Sports Illustrated.

A few sports and entertainment-focused outlets that dominated the trending searches in the related areas aside, it can be argued that Google’s algorithm favors U.S. news organizations that are considered by the public more or less reliable, credible and believable, especially by the voters who identify themselves as Democrats, as seen in the higher visibility of publishers like CNN, NYT and NPR.²⁹

The volume of each trending search in the U.S. varied highly with a minimum of 20,000 searches for the top 10 queries of the day. The peak was more than 10,000,000 searches for “Kamala Harris” in the aftermath of the presidential elections and “Sean Connery” after the actor’s death. Google picked NPR as the top source for the Harris story and BBC News for Connery.

It was also noted that the top results in 94 of the 541 trending searches featured local media outlets such as Des Moines Register and the Houston Chronicle. Apart from sports- and entertainment-related queries, the highest volume trending search that Google picked a local news outlet as the top result was “Who is the president of the United States.” More than 500,000 queried this question on November 8 and Google picked Baltimore Sun as the top result.

While this simple study for Google’s trending search results in the U.S. provides us with a rather basic insight for a comparative approach, we collected the Turkey data much more meticulously as described in the methodology section.

Out of the 84,430 queries in our monitoring period (October 5-November 5, 2020), 63,735 of them were on Google’s search engine as we recorded SERPs³⁰ for the two high-volume Turkish news keywords we selected and the 10 trending queries of the day as listed by Google Trends.³¹

In the “son dakika” (breaking news) and “haberler” (news) queries, we counted the top three slots in Top Stories carousel³² for this analysis. Google’s algorithm reserved a total of 4,488 top slots in the Top Stories carousel to only 17 Turkish media outlets in these queries.

3,811 of these 4,488 slots (85%) were reserved to three pro-government media outlets: Hürriyet, Sabah, and Milliyet.  Only the following 9 outlets appeared more than 10 times for each (Fanatik and Fotomaç are sports newspapers):

Considering all results in this count, it was seen that Google gave 90.6% of the top slots in its highest volume news-related searches to the Group B outlets, which include highly partisan media companies. The independent outlets in Group A were given only 8% of the slots. Apart from Sözcü, which was given 356 slots, there were just two more Group A outlets in the list: Halk TV and Yeniçağ, which were given by Google only one top slot in thousands of news-related queries. While other types of sources, including aggregators and sports websites, were given 1.4% among the top slots, no local media outlet was given any at all. The figures of slots by media groups are below:

Google News is an application and website that attracts many avid news readers and drives significant traffic to publishers. In Turkey, too, media outlets try to be more visible on Google News by using various methods, including prominently displaying its widget on their websites.

In order to find which outlets are more visible on Google News, we monitored its home page³³ in the same period and every hour, recording each media outlet and their content featured by Google in the five main boxes. As a result, we collected 20,695 slots’ data, including details about the title of the featured news story, its URL destination, and page position.

These 20,695 Google News slots were given to 182 news sources, which seem like a more diverse base compared to the news-related result pages on Google’s core search engine. When the details are considered, though, it is seen that diversity is a problem on this count as well. Out of these 182 news sources, 6 of them got more than 1,000 slots each, having more visibility than the rest of the 178 outlets combined.

Google News provides pro-government outlets in Group B, alongside their sister companies, significantly higher visibility and traffic, which is disproportionate to their overall digital footprint and social media engagement. Only Sözcü, Cumhuriyet and Yeniçağ from Group A were in the list. No outlet from Group C, again, and the “D” in the graph below groups all other sources like news aggregators:

Compared to the Google search engine, independent outlets in Group A are given more visibility by Google News. Still, the pro-government media outlets in Group B, including those with a heavy political bias, received 73.8% of the Google News “pie” against the independent media’s 26.2% visibility. The figures in the chart below denote the number of slots each group was provided on Google News during the observation period:

Meanwhile, the interests and preferences of the average news user in Turkey should not be a mystery for Google, as it is the dominant information gatekeeper that routinely collects and analyzes vast amounts of news-related consumer data. As seen in the graph below, Google Trends data reflects that independent outlet Sözcü regularly attracts considerably more user interest compared to the pro-government flagship Hürriyet, while user interest for independent outlet Cumhuriyet is comparable to the pro-government heavyweight Sabah.

 

All these figures demonstrate that Google provides Turkey’s independent outlets in Group A with significantly lower visibility, despite their higher quality content with deeper public trust, broad digital reach, robust social media interactions, and faster growth compared to the pro-government outlets in Group B.  And Google’s algorithm in Turkey still channels disproportionally more users to these pro-government newspapers while suppressing the independent outlets in contrary to the popular demand reflected in its own search trends data.

One may question the reasoning behind Google’s algorithmic preferences in Turkey. Why would a U.S.-based digital giant favor a set of media outlets known for their partisan news coverage over another set of outlets with more balanced/fair journalism, despite the higher user demand for the latter group’s content and brands?

The short answer is we don’t know. Google’s algorithms are not public and the company has yet to answer why it does not enforce in Turkey the content quality and anti-disinformation practices that it enforces in other countries like the U.S. and the UK.

Surveys and Interviews

Clicks alone, whether they come through Google searches or social media interactions, cannot determine a news article’s value or a media outlet’s public impact and democratic contribution.³⁴ Considering the political, economic and technological challenges that Turkey’s independent journalists face, we sought to support our quantitative data-based findings with qualitative information by talking with them about their attitudes and needs. We focused on organization, human resources, operational strategies, distribution, finance, and sustainability.

The main findings of our survey can be summarized as follows:

• Organization: In Turkey, the typical independent media outlet on the national level has 17 employees, while this number drops to 7 on the local level. On average, 85% of employees are in editorial teams, while the rest operate in other areas including administration, technical, and marketing departments.
• Profitability: 60% of national outlets report that they were profitable in 2019, while 50% expect profits in 2020, too. Meanwhile, only 2 out of 7 local media outlets reported profits in 2019, and only these two expect profits in 2020, too, despite falling revenues due to the pandemic.
• Sources of revenue: 60% of national outlets’ revenues are sourced in advertising, while the share of reader revenues is 13%. Other revenue sources, which comprise 27%, include grants, funds, and equity capital. In local media, on the other hand, all revenues come from advertising.
• Sources of expenses: On average, 70% of costs are personnel-related expenditure (salaries, etc.), while technological/technical costs (coding, website maintenance, etc.) are the second biggest category with 18% of the share.
• Sources of traffic: According to their executives’ self-reported figures, Google has a share of 35.7% and social media platforms have a share of 32.4% as the main sources of digital traffic to national independent media. Among social media platforms, Twitter has a 63% share and Facebook is at 31%. When it comes to local media outlets, Google’s share drops to 21.5% while social media’s share rose to 36%. Social media platforms’ share also changes in local media with Facebook becoming the main source of traffic with a whopping 94%.
• Reader engagement: 26% of the users of independent national media outlets -and 33% of the local outlets- can be described as “loyal readers” who visit their websites much more frequently than the fly-bys. Considering that this figure is lower than 10% in Turkey’s largest media outlets that are now under the influence of the government, it can be safely argued that the media has a greater potential to establish sustainable business models supported by reader revenues. Their current capacity in this regard, however, is quite limited, as the national outlets only have an average of 780 digital subscribers or paid members. In local media, this figure drops to just 20.
• Differentiation of journalistic product and services: Roughly half of the independent national outlets and only 14% of local publishers use New Media formats, including podcasts and newsletters. No such new product or format had an immediate significant benefit, they say. They also lack adequate human resources and expertise to differentiate their journalistic product and services. For instance, most of the companies in national and local media have no dedicated video, product, or marketing teams. As another example on the distribution side, only 16% of national and 28% of local media outlets say they have at least one dedicated employee for search engine optimization or social media operations.
• Most popular digital tools: When asked which digital tools and programs that their teams use most, the most frequent answers were various types of (paid and free) software produced by Google, Facebook, Apple, and Adobe. Other frequently-mentioned ones were WordPress, Hootsuite, Crello, and Da Vinci Resolve, as well as several video production, distribution, and search engine optimization tools.

• Journalists are seen as the strongest HR asset: As seen in percentages in the graphs, a vast majority of executives in both national and local independent media consider their editorial teams, including journalists, as their strongest asset in human resources, although they still would like to improve and expand these teams to increase their journalistic capacity.
• Product development is seen as the weakest link: Product and software teams are widely seen as the weakest link in a typical independent media organization, followed by marketing and public relations.
• Ad and reader revenue in the focus of news business: Independent media executives attribute equal importance to the need to increase reader revenues and ad revenues. Other sources that can further diversify the revenue stream, such as grants, received more mixed comments.
• Most outlets see themselves as unsustainable: Only around half of the independent outlets at the national level consider themselves as sustainable businesses with an established strategy that they would continue in 2021. On the local level, this figure dropped to one-third, as most media executives said they were not sure whether they are currently sustainable. Interestingly, despite the bleaker outlook, most said that they would continue with their current business strategy.
• Platforms are seen as undermining quality journalism: A vast majority of independent media executives think that social media platforms and search engines fail in serving their quality news content to users. National media outlets complain more about these distribution platforms, compared to the local media, and social platforms scored worse than search engines in this regard.

To better understand the “hows” and “whys” behind all these “whats” affecting the capacity and impact of Turkey’s independent media, we conducted in-depth interviews with the same journalists and managers following the survey.

As summarized by Medyascope’s founder Ruşen Çakır, an award-winning senior journalist, Turkey’s independent media “may not be presenting a great performance at the moment, but it is mostly related to the limited resources they currently have.” According to Çakır, “political powerholders try to make sure that we continue to have limited resources” and in spite of the challenges, the independent media still survives as “it did not get crushed under the ongoing shocks.”

Glimpses of hope for sustainable news businesses developed by independent journalists are everywhere:

• Sözcü has a size and digital reach comparable —if not higher— than the biggest pro-government outlets.
• BirGün has already got 5,000 digital subscribers. It is remarkable, as independent outlets had got virtually zero digital contributors just two years ago.
• Gazete Duvar is among the leading independent media outlets in stepping towards revenue diversification, as 30% of its revenue now comes from ads, 20% from readers, and 50% from funds.
• Evrensel, one of the oldest news websites in Turkey, is still innovating by sustaining a smooth user experience, and by experimenting in new formats and platforms, like TikTok.

Even more “old-school” publishers like daily Cumhuriyet are currently experimenting with New Media forms, including podcasts. They have planned to move away from “news agency journalism” and started to train their employees as the new focus is exclusive stories. Meanwhile, other newspapers like Karar venture into more platforms like YouTube. Replica editions, or e-papers, also start to bring considerable digital revenue to some legacy publishers.

Political and technological challenges were combined with the difficulties related to COVID-19 in 2020, forcing many media outlets to take hard decisions. For instance, the Diken news website plans to shut down its physical Office permanently. However, the pandemic may also have some positive transformational effect, as the media executives we spoke to said that they are now more aware of the importance of revenue diversification as advertising revenues fall. The pandemic-driven remote working practices also benefited the media outlets in some other ways, such as by significantly reducing office and transportation costs.³⁵

For 2021, all independent outlets said they want to engage with their readers better through exclusive journalism with added public value, but they have three main concerns on their minds. The first two can be summarized as follows:

• Financial resources remain under heavy strain. Many outlets seek to hire more journalists and supporting employees to expand their news businesses, but they lack the funds needed to do it. Perhaps more importantly, most news executives are not able to draw up a clear strategy for financial sustainability, as they cite various political, economic, and technological externalities and uncertainties for 2021 and beyond. It reaffirms the urgent need to better support the independent media’s abilities to become financially sustainable.
• Limitations in human resources are among the main factors that impede the growth of independent media. There are two dimensions here:
  • Although there are thousands of unemployed journalists in Turkey, the human resources pool for reporters and editors skilled enough to succeed in the modern news business is quite shallow. Teyit, for instance, has to train (or re-train) its own fact-checkers because of this deficiency. And as Halk TV’s Yeşilçınar noted, “most of those big news stories are actually published by interns with almost no editorial oversight.” As a result, most news websites including many in the independent media become mere aggregators of social media or engage in “copy-paste reporting” with little or no public value. Hence, capacity building and re-training of journalists with a focus on digital skills remains as key activities to support independent journalism in Turkey.
  • There is also an ever-shrinking human resources pool for supporting roles in the news business. For instance, it is currently extremely hard for a smaller outlet to hire a full-time front-end or back-end developer. Many software engineers have moved from Turkey and those who remain are now “too expensive” for a small media outlet, as one participant said. And when a news outlet works with third-party providers for such services that closely affect their journalistic product, the result becomes usually unsatisfactory because such providers are not their employees who internalized their journalism. As Teyit’s Foça said, “mediating mechanisms should be established” to link independent media outlets with a skilled and accessible pool of human resources, especially for technical roles.

The third concern of the editors of Turkey’s independent media is the largely negative role of digital platforms, particularly Google and Facebook. Here are some of their comments:

• Biggest platforms incentivize clickbait in Turkey: “Google has an awful impact on Turkey’s digital content ecosystem. Perhaps it is unintentional but they have become a source of manipulation that corrupts all digital metrics in the ‘news and media’ category. As a result, we have arrived at a point where we can’t produce refined competition data. It should be added that most social media platforms are also too open to clickbait and manipulation. However, these platforms don’t bring your website high quality, engaged users.” — Sözcü CDO Erhan Acar
• Political censorship affects search indexing: “Our news website, Oda TV, has been shut down [by Turkish authorities] seven times. We are currently serving a mirror website at odatv4.com. Whenever we changed our domain after each shutdown, we lost our Google index and search ranks. Such drops in search rankings bring significant economic blows as our readers cannot even find our current website. Google should develop a policy for countries like Turkey [where news websites are arbitrarily shut down with political motives].” — Oda TV editor Barış Terkoğlu
• Google favors big players: “Well–functioning of Google’s policies in Turkey is one of our three most important needs because search is the source of a significant portion of our digital traffic and revenue. Unfortunately, Google is currently penalizing smaller websites while ignoring the violations of the bigger news websites. Because of their high potential for traffic and advertising, Google privileges them. It means that the small websites are likely to stay small and the biggest ones to stay big. [Editors’ note: As the playing field is not level on Google’s search engine, newcomers in the Turkish news business rely more on other channels, including direct traffic and social media]. We also know that the bigger news websites systematically file complaints to Google to suppress emerging competitors. These smaller sites get penalized by Google, while bigger ones are not sanctioned, and we know that this whole mechanism turns into a sort of market in Turkey. Sometimes political reasons are also involved.” — Karar editor Fuat Atik
• Multiple respondents stressed that Google should activate the anti-clickbait policies that it practices in other countries. BirGün editor Hakan Demir also pointed out the fact that the requirement to employ “expensive” SEO experts gives the bigger news organizations an unfair advantage over the smaller ones, depriving them of organic search traffic.
• YouTube’s indifference moves away audiences: Halk TV’s Yeşilçınar said that their live broadcasts on YouTube, which frequently draw 10,000-20,000 viewers at a time, often gets disconnected and YouTube automatically channel these live viewers to their competitor Habertürk’s broadcast. Even after they changed their technical infrastructure, including encoders, the problem persisted and the Google-owned platform failed to respond or help to solve the problem. Çakır, on the other hand, complained that YouTube sometimes blocks Medyascope’s news videos but they cannot contact anyone to object to the decision. “Most of the time, YouTube’s decision is so sloppy that if we could explain ourselves, they would remove the ban immediately,” he explained.
• Facebook algorithm favors ‘tricksters and trolls’: Most media executives think that Facebook’s algorithm downgrades quality journalism, especially when it directs the user to their news websites while favoring the “tricksters and trolls,” in the wording of Oda TV’s Terkoğlu. Moreover, Facebook took down several pages owned by independent media and journalists in recent years even without a court order or a government complaint. “Abandoning of chronological timelines [in social media feeds, replaced by algorithmic curation] has had a remarkable negative impact,” Evrensel editor-in-chief Fatih Polat said. “Facebook is suppressing our reach,” Terkoğlu added. Several other participants call on Facebook to drive traffic away from “click baiters and copy-pasters” while increasing the visibility of “trusted journalism outlets” in its news feed.
• Twitter is not doing enough for news monetization: Independent media outlets voice relatively more positive opinions about Twitter. Most of them use the platform to engage with the most ardent news consumers. For Halk TV, Twitter is also a source of significant income thanks to the “video cards” feature. This is not the case for all publishers, though. The video-oriented Medyascope’s Çakır, for instance, complained: “Twitter brings no benefit to us. There is no revenue from Periscope. There is no efficient way to monetize your news content with ads. And the Twitter app that enables the audience to donate to the publisher during a broadcast is not operational in Turkey at the moment.”
• Turkish platform Dergilik’s double-standard for local media: The Turkish telecom corporation Turkcell’s Dergilik application is the leading platform for replica editions of hundreds of newspapers and magazines, like Apple News+ in many other countries. For local newspapers and independent/individual publishers, it is even more important to reach a large audience. The Dokuz Eylül newspaper’s editor-in-chief Murat Attila, however, complained that Dergilik declined to let them use the paid subscription feature. “We were ready to pay Dergilik a 40% commission fee for each of our subscribers. They rejected us, but the Yeni Asır newspaper was accepted in their system,” he said. Yeni Asır is a pro-government daily in İzmir where Dokuz Eylül is also based.

Despite largely negative opinions about the role digital platforms play in supporting quality journalism in Turkey, independent media executives admit that they have a symbiotic relationship with these companies. “Although I criticize it a lot, Google’s contribution in expanding our reach is substantial. Moreover, I actually prefer digital ad revenue over reader revenue and grants, because getting your money from thousands of advertisers makes you editorially more independent than relying on a smaller number of readers or funders who may have certain expectations in your news content,” Terkoğlu explained.

Turkey’s local media outlets, including the ones that are recently funded by Google or participated in Facebook’s journalism workshops, have been hit by all these negative impacts. The government-controlled Press Advertising Council’s (BİK) decisions to impose fines on independent outlets³⁶, as well as various types of local pressures, cornered these outlets in 2020. Hundreds of local newspapers have been shut down. As a result, one of the main problems that local media executives cited was the lack of competition. “If there was another outlet that can challenge us and compete with us effectively, we could be better at improving ourselves,” Tigris Haber’s Ali Yılmaz said.

As others point to the dangers of being dependent on ad revenues, the local news website Manavgat Haber’s Adem Polat claimed that they are planning to overcome the problems by producing original news content and switching to a reader-first revenue model in 2021. “We rejected some grant offers because they had wanted to use us for their propaganda. We were not profitable in 2020 and we don’t expect to be in 2021, either. We will be profitable in 5-6 years,” he added.

Individual journalists have also recently started to operate as independent news outlets in various mediums. Senior journalist Murat Yetkin, for instance, founded his own news website, YetkinReport, after he resigned as Hürriyet Daily News’ editor-in-chief in 2018, following the company’s takeover by Demirören. Speaking to Duvar English editor-in-chief Cansu Çamlıbel, Yetkin said it was a “necessity” for him to establish his own website to “remain independent.”³⁷

Yetkin recalled how Turkey’s biggest news organizations could not even report the shocking resignation of Treasury and Finance Minister Berat Albayrak in November 2020 because they were waiting for a green-light from the government to break the news. “The outlets that were described as the mainstream in the past are now poisoned soil. Perhaps there are still a few wildflowers that manage to blossom there, but they are exceptions, as the rule is that the soil is poisoned. Independent journalism is not possible there anymore,” he said.

He also admitted that he could afford to work on his own. “I have a pension, my kids are grown up, and I own my house. I also don’t have any lavish spendings. As a result, I had the luxury to engage in independent reporting by myself,” he explained. Currently, Yetkin Report “can only pay its expenses” as Yetkin hopes to establish a newsroom, hire reporters and technical personnel, and pay royalty fees to the contributing authors one day. According to Çamlıbel, the Ankara-based Yetkin Report is already displaying the positive qualities of ‘old Turkey’s mainstream media as it strives to reach a diverse audience with news content enriched by multiple sources on various sides of the political spectrum.

New Media forms are also used by individual publishers who once worked for big media outlets. Journalists Özgür Mumcu and Eray Özer’s “Yeni Haller,” for instance, is a popular podcast. Mumcu, the son of the legendary investigative journalist Uğur Mumcu, said he had been writing opinion articles on daily political issues as a Cumhuriyet columnist. His new podcast, however, is more focused on evergreen issues of culture. “Journalism podcasts are taking baby steps now, but I believe that we will see a trend in Turker similar to other countries [where journalism podcasts are highly popular]. There is a potential for rapid expansion,” Mumcu added, stressing that “Yeni Haller” already attracted an “educated, highly interested and engaged audience.” He also said that the two journalists have to deal with all aspects of the production, which provided them a “priceless autonomy.”

The newsletter is another popular New Media form that attracts Turkey’s independent journalists who fled the politically-invaded, large-scale news organizations. Kapsül became one of the most popular daily newsletters on Turkish news soon after it was founded in 2020. “Turkey’s internet media is doing what can be described as ‘fast food newsmaking.’ Most outlets copy-paste each other’s content or following editorial policies burdened by their ideological baggage,” Kapsül editor Minez Bayülgen said.

Kapsül, still in its first year, currently curates news, but it aims to produce original reporting soon. “Kapsül is now like a filter between the pro-government and the opposition media. We scan Turkey’s media, eliminate commentary and propaganda, and deliver the facts in news stories to our readers,” she added. “We indeed wish to produce our features and investigations because there is a vacuum in this barren and low-quality environment in today’s Turkey.”

 

---

Conclusion and Recommendations

Only four years ago, Turkey’s “old” mainstream media outlets all dominated the Comscore and Gemius reports that ranked the country’s most popular news websites objectively. As most of these outlets come under direct or indirect government control, several newcomers emerged and many of them are already in the top league in digital reach and social engagement. In Gürsel’s words, at least some of these outlets seem to have “the kernel of a new ‘mainstream’ media.”

Yes, “the kernel” reflects the potential and it is obviously there. However, these independent outlets need help to become the new mainstream, which retains the positive qualities of the old mainstream while eliminating the negative ones. It is not an easy task, though. Dozens of journalists are in jail for doing their jobs, while thousands are unemployed or are stuck in self-censorship. More journalists are physically assaulted.³⁸ Digital transformation and the pandemic brought even heavier economic pressures on them. Moreover, many independent outlets and journalists also inherited some of the inadequacies and errors of the “old mainstream.”

As Turkey’s independent news outlets need help on their way to become “the new mainstream” with financial sustainability and democratic prowess, our recommendations for all stakeholders, including the donors, are based on three articles of the IPI’s 2021 Action Plan:

1. Connecting (and uniting) journalism: Turkey’s independent media landscape is too fragmented. Those who are committed to quality journalism should unite and combine their forces. Such a nationwide coalition of journalism would not only share digital newsmaking skills and resources with each other for more effectiveness and efficiency but can also engage with digital giants and the government as an alliance to demand fair access to and treatment by digital platforms.
2. Understanding (and overcoming) key challenges: Turkey’s independent media has two difficult problems: Monetization and human resources. Supporting the quality journalism ecosystem, in the long run, is much more important than providing short-term funding to a specific outlet or a journalist. To succeed in the evolving business and technology of modern journalism, independent media outlets should have affordable access to full-stack consultancy groups, incubators, and accelerators. Third-party providers of services such as coding, product development, and customer relations should have expertise in the news media business. The new structures and parameters for effective resource-sharing and fair access to the services and products required for quality journalism should be determined through more dialogue and cooperation between all stakeholders. Capacity-building initiatives and financial support schemes should also remain a top priority to elevate the standards of journalism.
3. Promoting journalism and dispelling disinformation: Driven by commercial interests, search engines and social networks are gatekeeper platforms that tend to favor the status quo and cooperate with governments even when they repress fundamental rights, including free speech. Their effects on Turkey’s news media should be closely monitored, and the necessary actions should be taken to cope with their negative impacts on democratic processes. Press freedom advocates and civil society should develop new, more effective methods to expand the reach of quality journalism and to enable competition on a more level playing field for digital news businesses. Platforms, mainly through their algorithms, should treat news outlets transparently and with fairness, prioritizing the public interest in their decisions that affect the distribution of news content.
4. Principles and policies endorsed by the EU will impact regulatory regimes in candidate countries and beyond, following the Commission’s Action Plan Against Disinformation. It is essential therefore that the EU strikes the correct balance in ensuring a fair, open, transparent, and viable platform for independent media to operate.

With this report, we tried to provide a detailed picture of independent media’s impact, capacity, potential, and needs. We can gladly assist other press freedom advocates and media researchers to build on our findings and databases. Follow-up studies can focus on issues including news consumers’ habits or perceptions, platforms’ role, and which media outlets break echo chambers while others reinforce them in Turkey. Comparative studies can also be conducted to cross-check the problems and develop joint solutions for independent media outlets in other countries, too.

 

---

 

About the Authors

Emre Kızılkaya is the vice president of IPI’s National Committee in Turkey, and the project editor of Journo, a non-profit news platform supported by Turkey’s Journalists’ Union (TGS) and the EU. He worked at the Hürriyet newspaper for more than 15 years in various roles including managing editor. Kızılkaya was selected as a Knight Visiting Nieman Fellow at Harvard in 2019 and was awarded as the Digital Journalist of the Year in 2018 by the Journalists’ Association of Turkey (TGC). Having a master’s degree in journalism, he is currently conducting a PhD research on public trust in digital media at Galatasaray University.

Burak Ütücü graduated from Anadolu University’s Journalism Department. In 2015, he worked as an intern at Hürriyet newspaper, and two years later he was an intern editor at a London-based production company. Ütücü is currently a freelance journalist and the editor of the Turkish news website Observe Turkey.

 

---

About IPI

Founded in 1950, the International Press Institute (IPI) is a global network of editors, journalists, and media executives dedicated to furthering and safeguarding press freedom, promoting the free flow of news and information, and improving the practices of journalism.

The organization has been active in Turkey since the 1950s to elevate journalism standards and advocate for the freedom of the press. IPI Turkey’s former board chairs and members include prominent journalists like Abdi İpekçi, Sami Kohen, Hasan Cemal, Metin Toker, and Ferai Tınç. Turkish journalist Ahmet Emin Yalman was among the founding members of the IPI.

 

---

Further Reading and Footnotes

• A coalition of 11 international press freedom, journalism, and human rights groups, including the IPI, warned following a four-day mission to Turkey in October 2020 that the press freedom crisis in the country worsened as the government continued to capture news media outlets, ratified a social media law to clamp down on free speech and further undermined the independence of democratic institutions.
• Turkey’s ruling party has used the distribution of public funds and public tenders to achieve almost total control over media ownership. Investigative journalist Çiğdem Toker explained how this mentality “promotes corruption at all levels of the state” in an article for Turkey Dispatches, a special IPI series on press freedom and the realities of practicing journalism in the country.
• While the researchers of this report welcome state regulators’ efforts and raising public awareness in the U.S. and the EU to cope with the multi-faceted problems caused or worsened by digital platforms, they are highly skeptical about Ankara’s latest drive to regulate these corporations. For a brief history of the Turkish government’s crackdown on online speech, as well as its most recent moves targeting digital platforms, see Deniz Yüksel. “Turkey’s Government Wants Silicon Valley to Do Its Dirty Work.” Lawfareblog.com. 9 December 2020.
• After we completed the data collection for this research in December 2020, three U.S. academics reported the findings of their study in the Washington Post, writing that U.S. “local news outlets rarely make it to the first page of Google News results” and “well-funded national outlets get the most attention.” See: Sean Fischer, Kokil Jaidka and Yphtach Lelkes. “How Google is hurting local news.” Washington Post. 22 December 2020.

1. More on comparative figures about Hürriyet’s and New York Times’ digital reach and engagement (in Turkish): https://www.Hürriyet.com.tr/yazarlar/emre-kizilkaya/Hürriyet-new-york-timesin-2-dakika-onunde-40690976 
2. In fact, Reuters Institute’s reports had noted that “Turkey and Kenya are urban samples that will inflate figures” and “2018 figures for computer use were likely overstated due to an error in polling,” but neither of these brief statements explains why “CNN Türk online” was still wrongly ranked as Turkey’s top outlet in digital reach even in the 2020 report (see page 84-85): https://reutersinstitute.politics.ox.ac.uk/sites/default/files/2020-06/DNR_2020_FINAL.pdf 
3. More on working as a journalist on the night of the coup attempt: https://www.poynter.org/news-release/2016/in-turkey-it-seems-the-critical-press-is-defenseless-at-the-moment/
4. More on the Journo.com.tr project: https://niemanreports.org/articles/as-erDoğan-cracks-down-turkeys-independent-journalists-need-digital-skills-and-business-acumen/
5. A crowdsourced project titled “Networks of Dispossession” provides an interactive Graph Commons that maps the relations and links of several business groups vis-à-vis media outlets and large-scale public development projects in Turkey: https://graphcommons.com/graphs/09426282-c311-4b39-be2c-1bd4f93b5771?auto=true
6. More on the Demirörens in RSF Media Ownership Monitor: https://turkey.mom-rsf.org/en/owners/companies/detail/company/company/show/demiroeren-group/
7. More on Sabah’s chair Kalyoncu: https://turkey.mom-rsf.org/en/owners/companies/detail/company/company/show/kalyon-group/
8. More on Yeni Safak’s chair Ahmet Albayrak: https://turkey.mom-rsf.org/en/owners/individual-owners/detail/owner/owner/show/ahmet-albayrak/
9.  More on Aksam’s chair Zeki Yesildag: https://turkey.mom-rsf.org/en/owners/individual-owners/detail/owner/owner/show/zeki-yesildag/
10. More on the financing of the Beyaz Holding media outlets with the help of AKP-controlled municipalities (in Turkish): https://medyascope.tv/2019/09/23/kanal-7-ve-ulke-tvnin-sahibi-beyaz-holding-son-sekiz-yilda-ibbden-1-milyar-123-milyon-tllik-ihale-aldi/
11. More on the RTÜK fine targeting Habertürk for broadcasting opposition MP’s remarks: https://bianet.org/english/media/235392-media-authority-fines-haberturk-tv-over-opposition-mp-s-remarks-during-program  
12. More on Habertürk owner Turgay Ciner: https://turkey.mom-rsf.org/en/owners/individual-owners/detail/owner/owner/show/turgay-ciner/ 
13. More on Google’s Journalism Emergency Relief Fund: https://newsinitiative.withgoogle.com/journalism-emergency-relief-fund/recipients/ 
14. Gemius defines “real users” as “the number of Internet Users (visitors) in a given target group who visited (generated at least one page view) the selected node(s) in a specified time period. This indicator relates to the actual number of persons – not computers, cookies or IP addresses.” More on the methodology of Gemius Audience: https://audience.gemius.com/en/methodology/overview/ 
15.  Toff, Benjamin; Badrinathan, Sumitra; Mont’Alverne, Camila; Arguedas, Amy Ross; Fletcher, Richard; Nielsen, Rasmus Kleis. “What we think we know and what we want to know: perspectives on trust in news in a changing world.” Reuters Institute, 3 December 2020. https://reutersinstitute.politics.ox.ac.uk/what-we-think-we-know-and-what-we-want-know-perspectives-trust-news-changing-world
16. Google is the dominant search engine in Turkey with 84% market share as of August 2020: https://www.karar.com/yandex-turkiye-ofisini-kapatiyor-1586093
17. Makhortykh, Mykola; Urman, Aleksandra; Ulloa, Roberto. “How search engines disseminate information about COVID-19 and why they should do better.” Harvard Kennedy School Misinformation Review, 11 May 2020. https://misinforeview.hks.harvard.edu/article/how-search-engines-disseminate-information-about-covid-19-and-why-they-should-do-better/
18. According to the search engine analysis and optimization tool KWFinder, “haberler” (“news” in Turkish) is searched for 24.5 million times per month, while the figure for the “son dakika” (“breaking news” in Turkish) query is 26.2 million in Turkey. BuzzSumo, another tool used in the industry, provides almost identical figures for both queries in Turkish. According to BuzzSumo’s keyword alert tool, Turkish digital outlets published an average of 1,049 news stories with the “son dakika” phrase in their headlines every day in December 2020, hoping to appear among the top results on Google to attract these users to their websites.
19. For more on the negative effects of “incidental news exposure,” see: Barnidge, Matthew. “Testing the Inadvertency Hypothesis: Incidental News Exposure and Political Disagreement across Media Platforms.” Journalism. 21, no. 8. p. 1099–1118. August 2020
20. Atik, Acar, Yeşilçınar and Terkoğlu parted ways with their outlets after our interviews.
21. The Gemius dataset did not include data for Diken and this outlet’s representative declined to share the figures about its digital reach. If the volume of social media interactions and search engine visibility are taken as a benchmark, one can argue that Diken’s overall digital reach stands somewhere between Gazete Duvar’s and Evrensel’s reach.
22. To complement the comments in the Foreword, it can be noted that CNN Türk’s website reached 9,159,280 real users (15.64% of the total audience) in October 2020, falling behind the top five outlets in Group B and the top two outlets in Group A.
23. For the details: https://docs.google.com/spreadsheets/d/1WgnCKbIJJWGowOhV-Zmb_sa0qWGu0fIt36fSm8IzZBo/edit?usp=sharing
24. Facebook API has limits on data collection, which affected some of the figures in this section, as not all Facebook and Instagram interactions could be retrieved. As a result, the figures do not include Facebook interactions for Gazete Duvar, Medyascope and Halk TV.
25. We collected the data with NodeXL and visualized it with Gephi by using the ForceAtlas2 algorithm. Distances between nodes and clusters denote the interactive relationships of Twitter users, and not the editorial policy or ideologic stance of the media outlets.  We painted the red and blue shades just to make it easier for a viewer to identify Group A and Group B outlets. Otherwise, there is obviously no strict line dividing Twitter audiences as such because all communities are more or less connected with each other. For how the visualization algorithm works, see: Jacomy M, Venturini T, Heymann S, Bastian M. “ForceAtlas2, a Continuous Graph Layout Algorithm for Handy Network Visualization Designed for the Gephi Software.” PLOS ONE 9(6): e98679. 2014.
26. The word “homophily” was coined in the 1950s by U.S.-based sociologists Paul F. Lazarsfeld and Robert K. Merton in a study on friendship in a mixed-race housing project in Pittsburgh. This concept’s algorithmic aspect has become the focus of academic interest with the rise of digital communities and platforms. It is now widely accepted that algorithmic preferences of platforms strongly affect a digital citizen’s information diet. In 2020, a study in the U.S. and Germany concluded that YouTube’s recommendation algorithm fosters the creation of highly homophilous communities in these countries. See: Kaiser, Jonas, and Adrian Rauchfleisch. “Birds of a Feather Get Recommended Together: Algorithmic Homophily in YouTube’s Channel Recommendations in the United States and Germany.” Social Media + Society. October 2020.
27. Reuters Institute’s Digital News Report 2016: https://www.digitalnewsreport.org/survey/2016/ 
28. In this period, the newsletter was not delivered in four days, hence we did not include the related data here.
29. See “U.S. Media Polarization and the 2020 Election: A Nation Divided” by Pew Research Center: https://www.journalism.org/2020/01/24/u-s-media-polarization-and-the-2020-election-a-nation-divided/
30. As all these “hot searches” are closely associated with fresh news stories, Google’s algorithm brought the Top News carousel in almost all of such queries. Still, we recorded both the carousel results, as well as the list of organic results below it. We disregarded the “snippet” section which rarely appeared in some searches.
31. A news-related search subject was frequently linked to more than one key phrase. We selected the first keyword that Google Trends cited for a particular trending search.
32. The first three slots of this carousel are visible on top of the SERPs in PC and the first two of them are visible on mobile phone screens, which bring disproportionately more visitors to news websites when Google gives them these slots. Although there is no academic or industry research focusing on click-through rates of the Turkish news content in Google’s Top Stories carousel, earlier studies in the English-language content had indicated that the top three organic search results get 75.1% of all clicks. See: https://backlinko.com/google-ctr-stats 
33. Again, we used “always clean” PC and mobile browsers in default settings to avoid personalization of the news results.
34. Schmidt, Christine. “Clicks are an ‘unreliable seismograph’ for a news article’s value — here’s new research to back it up.” Nieman Lab. 14 February 2019. https://www.niemanlab.org/2019/02/clicks-are-an-unreliable-seismograph-for-a-news-articles-value-heres-new-research-to-back-it-up
35. For more on this issue, see the series of video interviews entitled “Journalism Under the Shadow of Pandemic” by İpek Yezdani from IPI Turkey: https://freeturkeyjournalists.ipi.media/journalism-under-the-shadow-of-pandemic-turkey-independent-media-as-a-basic-need/
36. “IPI condemns record ad ban on Turkey’s Evrensel newspaper. ” https://ipi.media/ipi-condemns-record-ad-ban-on-turkeys-evrensel-newspaper/
37. IPI Freedom Dialogues, December 29, 2020. https://freeturkeyjournalists.ipi.media/tr/ipi-ozgur-sohbetler-geleneksel-medyanin-yoklugunda-bireysel-cabalarla-surdurulen-bir-gazetecilik-oykusu-yetkin-report/
38. According to the main opposition MP Utku Çakırözer, a former board member of IPI Turkey, five journalists were physically assaulted in the first two weeks of 2021. This number was 17 for the whole year of 2020 and 34 for 2019. See: https://bianet.org/bianet/ifade-ozgurlugu/237685-yilin-ilk-15-gununde-bes-gazeteci-fiili-saldiriya-ugradi