Coordinated Campaign to Address MDH online in Ethiopia

      A. Overview of the project

 #Gira-Kegn, the Coordinated Campaign to Address Mis/Disinformation and Hate Speech online in Ethiopia, was bornout of concerns regarding the alarming rate of misinformation and hate speech spreading on digital platforms. While previous initiatives like ETH 052 focused on combating this issue through mainstream media, it became evident that MDH (Mis/Disinformation and Hate Speech) is primarily prevalent on social media. Thus, this project aims to confront the problem directly on social media platforms where MDH thrives.

Launching the campaign across various platforms was crucial not only to address the issue at its source but also to leverage the immediate data obtained from each post. This data provides valuable insights into reach, engagement, and potentially even behavioural change. Throughout the project implementation period, we engaged in a range of activities, including influencer screening and selection, brand development, training, campaign launch and promotion, content creation and distribution planning, and data analytics.Below are major highlights of key achievements in brief.

    B. Key achievements and outcomes

 Over the course of 8 weeks- content distribution period of the project, a number key achievements were observed. The following four are carefully selected examples:

Diverse & Coordinated Contents: each influencer created and distributed 5 contents, with additional 6 promo contents from us, and additional contents from Tikvah, more than 65 contents were created for this campaign. Although the contents are diverse in nature, they are coordinated through the #Gira-Kegn brand colour, logo and tagline as well as music. These contents, especially the videos can be repurposed for other distributions, if needed, say as ads or fillers on TV. All contents highlighted the need to pause, reflect, and look all sides before making judgments.

Reach & Engagement: the combined total reach of the contents created and distributed under #Gira-Kegn campaign reached 5 millions of people. The engagements through comments, shared and likes were encouraging also, with 302 Thousand Total Interactions* of which 287,000 were Likes, 8,100 were Comments, and 6348 were Shares. Going a bit deeper analysing the types of comments, the vast majority of impressions were very positive. The first launch week, for instance generated very supportive comments of  about 92%, while only about 7% of comments contents are negative. By all standards, these are very good figures for any similar intervention.

Mainstream Media Coverage: the reach of the #Gira-Kegn campaign was not limited to social media spaces. A number of mainstream media houses gave the campaign an extensive coverage, both as a news item and a program/ show input. 16 major media outlets amplified the initiative- VoA did a couple of shows on the project. EBC, Fana, OBN, ENA and ESAT gave an intense coverage to name a few of them. This helped boost the campaign and it is one incident where the new and the old media collaborated on a single cause.

 Influencers’ Growth: #Gira-Kegn brought some of the influencers an opportunity to grow both personally and professionally. As a result of this campaign, one of the influencers, Ye-Lij-Magna, became a judge on Ethiopian Idol show, to be televised on ETV. Influencers like Nati-Xia and Musse Solomon who regularly publish current affairs, news type contents became so equipped with fact checking tools, they made information verification their standard criteria in sharing a story. Almost all influencers follower base grow. The pre and post campaign follower and likes counts shows, there is a steady growth for all influencers involved.

 Pre and Post #Gira-Kegna Followers and Likes Increase

The Table below shows Influencers’ Followers base and likes increase in Pre and Post Campaign period.

       C. Challenges faced

 As any project, this activity is not free of challenges. The following three are carefully selected to highlight some of the difficulties and obstacles we face along the way:

Social Media Blockage: in February 2023, following the unrest within the Ethiopian Orthodox Church, the government blocked most social media platforms. People now can only access sites like Facebook, TikTok, YouTube and Telegram through VPN. This was a major challenge for this project. Trying to promote the campaign through paid ads for instance became impossible, as it is very difficult to geo-locate users for targeted ads. Generally it is also believed peoples’ time online would probably decrease as a result of the blockage.

 In-Kind Contribution Delay: the other operational challenge we faced was the super delay in Dexis’ Procurement team to open a bid & select a winner to carry out the branding and promotion activity. That caused about 2 months delay in launching the project. The activity was intended to be launched late January or early February, but we could only do it in early April.

 Personal Issues of Influencers: some of the influencers went through difficult times because of their contents. Abel Gebrekidan was imprisoned in relation to the problems at the  Ethiopian Orthodox Church. A few others were somewhat difficult to manage because of big super ego. A couple of influencers protested the appearance of government officials for the launch and tried to hide themselves. The delay in the launch made a few of them gang up and protest to ask for money before the completion of the project. As expected dealing with personalities with huge follower base was a considerable challenge.

   II. Implementation Highlights

 The project went through months long different phases of implementations. Below are major reflections and recollections of key implementation highlights:

Influencers Selection: One of the prime activities of the project was to develop social media influencers, and online media houses selection criteria that constitute reach (number of followers), engagement rate (relation with audience), content types (we aspired for diverse content types), reputation (to avoid those known for MDH and controversy), and also ease of doing business with. With these set pf criteria, we shortlisted about 50 influencers, had a phase to phase with all of them in group and individually in attempt to still in the idea in them, and also to gauge their interest. Out of them, we trained 14 influencers, with a view of producing 70 contents, but three of them dropped out after training, mainly because of busy schedules, and we implemented the project with 11 influencers- two of them digital media houses- Dire Tube and TikVah, while the rest are individual influencers.

Brand Development: in order to identify the contents of our campaign from regular contents of the influencers, we set out to make our intervention branded, and we spend a considerable amount of time in development process. After going through various choices, we, in agreement with involved influencers eventually settled with #Gira-Kegn, which literally means “Left-and-Right”, implying the need to look in both or all directions. With creative logo templates, tagline, and original intro and background music, as well as brand usage guide, the branding work was well executed. We encountered one minor coincidental branding clash 12 days after launching the campaign. The state-owned ETV started a drama series with the same name Gira-Kegn, featuring very famous Ethiopian actors. This is one of the problems of using generic names as brand titles. Nonetheless, we cannot be sure of the impact of the drama on our campaign.

Training: we trained the influencers for two days on key issues of what mis/ dis/ mal info is, how one can identify and counter them, how content creation can be used as mitigation for disinformation, with open debates and discussions. We provided trainees with lots of reading materials and training slides as well as fact checking tools such as Google Reverse Image, Tiny Eye and the like. Some of them really picked it up well, and used these tools not only for contents for this campaign but for others as well. The training was so inspiring and eye-opening for most of them, and helped them create valuable contents.

 Content Creation: content creation against disinformation and in favour of peace, dialogue and tolerance was the guide given to them. Most of them did really well- creating factual, informative and educative contents, with clear examples on display to strengthen their messages across. With very few exceptions, may be a couple of influencers, the rest approached the topic with serious tone, making it lack a bit of entertaining flair. As the launch date delayed because of late in-kind contribution procurement process, we ensured each content creator finish all five pieces before launch, giving us enough time to review. We didn’t censor contents. Only once, we had to make corrections with one of the contents which unintentionally showed how to fish accounts on Facebook. The fact that contents were completed before launch made them somewhat timeless, that can be repurposed repeatedly for various platforms.

 Launch: as the launch delayed was in 2 months, we intentionally made it coincide with the World Fact Checking Day observed on April 2^nd, following April the Fool’s Day. It was for the first time that Fact Checking Day was commemorated in Ethiopia, and it attracted huge media attention. The launch party was attended by senior government officials, academics, members of the press and diplomatic representatives from the US Embassy. It was a successful event.

Media Amplification: as we were deprived of the chance to promote the campaign online or on social media, because of the blockage in February, we invested heavily on mainstream media to cover the initiative. The following are key coverages given to the campaign with their online links:

1. ETV- https://www.youtube.com/watch?v=QcCiqp3s8qM
2. Fana TV- Kelemat Show-https://youtu.be/YcBdVZpk_xY

3. VoA- https://fb.watch/jROYpcy7Jn/
4. ENA- https://www.facebook.com/1490190331223441/posts/3449670295275425/?mibextid=ou2O6MZ8of4BGb7x

https://youtu.be/vgyMTyveahc

5. Ahsam TV- https://www.youtube.com/channel/UCNhq
6. Hamer Broadcasting Service- HBS- https://www.youtube.com/watch?v=WV5dZB7FYo0
7. Addis Maleda
8. Ghion Media
9. TikVah-
10. DireTube-
11. Fana Radio- (LIVE phone in)
12. Bisrat FM- Hulu Addis Show- (LIVE phone in)
13. Harambe Radio-
14. Capital Newspaper-
15. ESAT

Content Distribution: content distribution calendar was shared with all participating influencers, in which, all six days of the week, except Sundays were included for sharing contents online. Most influencers follow the calendar strictly, a couple of them needed to be reminded, namely, Ab bella on Facebook, and Tikvah on Telegram. The rest did really well. DireTube, loved the approach and over delivered with seven #Gira-Kegn branded contents requesting the audience to discuss on key issues of concern. The page continued to deploy same approach on critical current affairs issues, even after the completion this project. All in all about 65 contents including promotional pieces were produced and distributed on all three platforms- TikTok, Facebook and Telegram.

    III. Data Analytics and Performance Monitoring

     A. Content Analysis

The contents created for the campaign can generally be categorized in to three types: Mis/Dis Info oriented, hate speech oriented and discussion questions (especially on Facebook by Dire Tube)

While Misinformation posts dominated content for the campaign (more than 80%), users engagement increased (a third of all comments) for issues dealing with Hate Speech.

       B.Reach and engagement metrics of influencers’ content

The campaign overall has generated 5 Million Views, and 302 Thousand Total Interactions* of which 287,000 were Likes, 8,100 were Comments, and 6348 were Shares.

There were a total of 65 contents made for this campaign, of which 8 of them are promotional pieces, while 58 contents are created by the influencers.

The posts made in the first 4 weeks passed 1 million views overall but the last period decreased by nearly a tenth as there were fewer posts.

TikTok was the mostly widely used platform for this campaign, covering nearly 80% of all views.

Dr.Seife and Musse Solomon were the campaign’s highest-performing channels, with nearly 50% of total views covered by their channel.

   C. Audience feedback analysis

As the content analysis shows, the majority of posts of this campaign focused on informing the public about misinformation, followed by hate speech countering contents and questions for discussions. The audience feedback to these contents show an interesting aspect that although Posts on Hate Speech account for only 12% of posts, more than a third (34.65%) of comments were made for these posts. Though it is hard to make a conclusion, given the small sample size, it is interesting to note that commenters on average were three times more likely to post comments on posts dealing with Hate Speech. This indicates how sensitive and triggering the topic is for social media users.

This discrepancy is clear when we observe that the “Like” and “Share” total percentage share is relatively roughly proportional to the number of posts.

Dr.Seife, Musse and Aster were the leading social influencers in terms of total interactions with their audience.

       D.Measurement of the Top 10 Performing Posts

 The top ten posts for the campaign account for nearly 80% of all views. They have an engagement rate of 5% which is twice as much as the average for the campaign.

Looking at the most well performing contents, 4 of them were produced by Dr. Seife, 3 by Musse Solomon and 2 by Aster- the top 3 well performing influencers. Dr. Seife led the metrics in terms of views and shares. Musse Solomon garnered the most Likes and Aster had the most comments by users.

What Makes their contents different from the others?

Dr. Seife produces health related contents, and there are interesting results even between his contents- his post on Fake News on Covid, for instance get 0 comments, (probably because it is somewhat outdated now?), while his other posts generated hundreds of comments- for instance on fake medical products (996 comments), a chiropractor who claims to heal 60 diseases and who was featured on ETV generated 639 comments as well as a defamation campaign on a juice product. The audience reaction may imply the big interest for health related contents.

Musse Solomon normally shares news, and current information but on this campaign most of his well performing posts were technology related- using AI to generate fake images, phishing attacks to steal information and his content on photoshopped image were well received by the audience probably implying the interest for technology related contents.

Aster’s contents deal with the sensitive issue of hate speech. As seen in the audience interaction of the overall contents of this campaign, hate speech sparks reaction. Contents that even call for unity and harmony are not free of polarized opinions. This area may need further investigation, but clearly shows the sensitive nature of the matter. Although hate speech countering contents account only for about 12.5 % of the overall contents, but they generated more than 34% of comments, nearly one-third of total comments.

        IV.Lessons Learned

         A. Strengths and achievements of the project

 #Gira-Kegn as the first coordinated online campaign to address MDH in Ethiopia has so many strengths and achievements, starting from being the pioneer of such initiatives. More specifically, the following can be highlighted as major accomplishments:

1. Media Attention: the campaign attracted huge media attention both on and offline, in mainstream and digital media with a wide coverage of the lunch event, the objectives and the need for fact-checking and information verification.
2. Public Attention: the huge media attention helped to mobilize large public interest on the issue of the campaign. As highlighted in the data section 5 million people, either watched or read contents created for this campaign. It has a huge reach in spite of the social media blockage.
3. Sparked Discussion: beyond reaching vast audience group, the campaign sparked intense discussion and at times even debates on the issues of MDH, especially on contents relating to hate speech. Although majority of the contents were dedicated to Mis/ Dis Info (more than 80%), third of the comments were generated by hate-related posts, which account for only 12% of contents, highlighting how polarizing these issues can be.
4. Adaptable Approach: the approach of coordinating social media influencers and online media houses can be replicated in other topic areas beyond MDH. On a preliminary survey in medical/ public health content creation for instance, we identified 16 medical doctors who are also TikTokers with a combined followers’ base of more 2 million people. They can easily be deployed for public health campaigns. The approach can work in other areas like entrepreneurship/ job creation, volunteerism, technology education, etc as well.

      B.Weakness and problems encountered

 Some of the weaknesses of this initiative emanate from the nature of project. This was not an ongoing continual program, and as a result was limited in scope, reach and influence, which is probably well exhibited in 3 major weaknesses identified as problems:

1. Short project activation Period: the project was run only for 8 weeks, and there are chances the issue could be forgotten or given lesser attention in times to come. Is it better to have a daily content for 8 weeks or say a weekly content for a year? That is the kind of question we need to reflect on.
2. Small Number of Influencers: we only worked with 11 influencers, in the universe of hundreds of influential people with millions of followers. Previous data from Opian advices to have the maximum impact on line in Ethiopia, coordinating about 130 influential figures can make one reach 30 million people, which can really be game changer.
3. External Factors: there are lots of things beyond our control such as the sudden & unexpected blockage, the political class, especially that of the government’s lack commitment to tackle MDH, etc are other areas of obstacles we encountered. A case in point can be the imprisonment of one of our influencers- detained for what he wrote, and released without charge. Such ongoing hardships will definitely affect the space.

      C. Recommendations for improvement

 There are a number of diverse ways to improve the campaign, if we are to do it again, or to extend to other content and topic areas. The following are some of the lessons we took in order to even improve the approach and its impacts.

Collaborative Partnerships: Establishing partnerships with relevant organizations, institutions, and influencers can enhance the campaign’s effectiveness. By leveraging the expertise, resources, and networks of these partners, campaigns can reach a broader audience, tap into new perspectives, and create a more comprehensive impact. Forming some kind of network amongst entities working in media literacy or fact checking space can be the way forward.

Targeted Messaging: Tailoring campaign messaging to specific target audiences can improve engagement and resonance. Conducting audience research and understanding their unique needs, concerns, and communication preferences can help develop targeted messages that effectively convey the campaign’s goals and resonate with the intended audience.

Monitoring and Evaluation: Implementing a robust system for monitoring and evaluating the campaign’s progress is essential. Regularly collecting data on key performance indicators, such as media coverage, public engagement, and reach, allows for informed decision-making, identifies areas for improvement, and helps demonstrate the campaign’s impact to stakeholders.

Influencer Training and Support: Providing ongoing and continual training and support to influencers involved in the campaign can enhance their effectiveness as ambassadors. Equipping influencers with the necessary knowledge, resources, and tools empowers them to deliver accurate information, engage with their followers effectively, and become trusted voices for the cause.

Call to Action: Including a clear call to action within the campaign can encourage active participation and support from the audience. Whether it’s signing a petition, donating to a cause, or sharing informative content, a well-defined call to action motivates individuals to take tangible steps towards the campaign’s objectives.

Community Engagement: Engaging with local communities and stakeholders can foster a sense of ownership and sustainability. By involving community leaders, grassroots organizations, and local influencers, campaigns can tap into existing networks, leverage local knowledge, and ensure the campaign’s relevance and long-term impact. Our campaign was only limited to online spaces.

Crisis Management: Developing a proactive crisis management plan is crucial to address unforeseen challenges effectively. Identifying potential risks, establishing communication protocols, and having contingency measures in place can help mitigate negative impacts, maintain the campaign’s credibility, and swiftly address any emerging issues.

Cultural Sensitivity: Recognizing and respecting the cultural nuances and sensitivities of the target audience is essential. Adapting the campaign’s messaging, visuals, and strategies to align with local cultural norms and values ensures that the campaign is relatable, inclusive, and avoids unintended offense.

     V. Sustainability and Future Directions

Sustaining donor-supported interventions is an all-time challenge for initiatives like these ones. We haven’t yet figured out the best business model say to fact checking practice in Ethiopia, probably the same is true to other similar African markets as well. Below are some grossly general ideas that can probably help sustain the project and possibly similar initiatives in the future.

   A. Strategies for sustaining the impact of the project

 Addressing MDH is an ongoing battle. With the proliferation of various AI tools like Deep Fake, the future is even scarily gloomy than bright. The following can probably help as future-oriented strategies:

1. Networking Fact Checkers and Organizations: a number of training initiatives took place in the last few years to alert media practitioners and social influencers to be vigilant of content they consume and create. A probable major missing point is the absence of networking among these bodies. Collaborative efforts can help avoid redundancy of efforts, wastage of resources and can also help gain lasting results.
2. Mainstreaming Fact Checking in Mainstream Media: no media in Ethiopia has a fact checking unit or desk or even a dedicated staff member for information verification and fact checking as a speciality. Most of known fact checking organizations grew out of major media houses- AP Check, AFP Check and the like. The same can be attempted locally.
3. Working in Collaboration with the Platforms: can be very critical in averting the spread and follow up effects of MDH. It is known that Facebook has been frequently criticized for its operations in Ethiopia and Myanmar fuelling ethnic and religious conflict. Omni through its affiliate Factify Ethiopia is now invited to be part of Meta’s Trusted Partners’ Program- privilege and support given to very few to help the platform take down harmful contents
4. Continual Media Literacy Initiatives: the ultimate strategy is to keep on doing what has been done through ongoing efforts, not only with focus on media houses but universities and journalism schools, and the public at large as well. This is a life-long engagement.

    B. Potential avenues for scaling up the initiative

 As mentioned above the approach of coordinating influential figures and digital media houses online can easily be scaled up in various content topic areas and life style preferences. The approach could probably work best in less political life affairs, very popular on social media spaces- such as health, wealth and relationships. By selecting key figures that can play prominent constructive roles on these areas, providing ongoing support and incentives, a change can be effected. A follow up activity of this campaign is MedConnect Ethiopia, a project under development to take this approach to the public health sector.

    C. Long-term work on Media Literacy awareness in Ethiopia

Raising media, information and technology literate content creators and consumers is the ultimate goal of initiatives like Gira-Kegn. This requires consorted efforts of media organizations, universities, donor agencies, professional societies, etc.  and it has to continue individually and collaboratively.

   VI. Conclusion

#Gira-Kegn, the Coordinated Campaign to Address Mis/Disinformation and Hate Speech Online in Ethiopia, stands out as a truly unique and innovative initiative, likely the first of its kind in the country. Through the collaboration of 10 influencers, the campaign successfully reached over 5 million people. However, challenges such as social media blockage and a lack of commitment from influential political actors, including the government, hindered the campaign’s potential to reach an even larger audience. Despite these obstacles, #Gira-Kegn remains a trailblazing effort in the face of Mis/Disinformation and Hate Speech in Ethiopia.

As the world embraces the proliferation of new and digital media, initiatives like #Gira-Kegn are likely to become the new norm. The lessons learned from this campaign have revealed the potential for launching similar interventions in other content and topic areas, such as Health and Technology. It is clear that social media-oriented interventions, where society is prominently engaged, are the way forward. #Gira-Kegn has pioneered this approach in Ethiopia, particularly in addressing the challenges of Mis/Disinformation and Hate Speech.

Moving forward, it is crucial to scale up and sustain these efforts in the ongoing battle against Disinformation and Hate Speech. The success of #Gira-Kegn provides a strong foundation to build upon, demonstrating the effectiveness of coordinated campaigns in combating the spread of harmful online content. By continuing to prioritize this issue and expanding the reach of such initiatives, Ethiopia can make significant strides in promoting a safer and more informed digital environment.