Written by: Annette Atieno, a dynamic communications professional and advocate for social change, with extensive experience in LGBTQ+ rights, strategic communications, and community engagement.

For those trying to shape public opinion in the fast-changing digital terrain, especially in settings as varied as Kenya’s, knowledge and navigation of online debates are critical. At the core of this work is the art and science of query design, which helps practitioners hone in on the information most important among the noise of the internet.

Understanding the Foundation of Query Design

Imagine you could walk up to a group of people and ask them a very specific question about what they think about you. And using these questions and how they responded to them, you could figure out how to navigate within the group thanks to their brutal honesty.

In digital analysis, query design is a meticulously structured process that involves crafting precise questions to extract meaningful data from vast online sources.  This process begins with a fundamental understanding of what you want to uncover, a step that might seem straightforward but is, in reality, a bit complex. Without a clear objective, even the most advanced tools can return results that are either overwhelming in volume or irrelevant.

We came together as feminist activists, digital researchers, and communications practitioners under the Global Narrative Hive and Komons. United by our shared concern over the growing influence of anti-rights actors, particularly in narratives surrounding family, gender, and sexuality, we designed a research approach that emphasized lived experiences and challenged traditional methods. Using an open-source digital analysis tool, Velma, we mapped and examined anti-rights actors and their tactics in real-time. Through regular reflection spaces and input from marginalized groups affected by these narratives, we developed a research objective focused on tracking how family values narratives influence online Kenyan discourses on sexuality, gender, and family. Our approach not only aimed to study these influences but also to create a living document of resilience and collective knowledge for future activist efforts.

Our query was carefully structured around key subtopics under the overarching theme of Family Values. We focused on two main subgroups: Normative Family Structure and Binary Gender Roles. Normative Family Structure refers to what a society considers the “normal” or “standard” family setup, which in many Western cultures has traditionally meant a married heterosexual couple living with their biological children. Binary Gender Roles describe the idea that there are only two distinct genders (male and female) with specific roles and behaviors assigned to each, such as women being caregivers and men being breadwinners. The team was tasked with identifying keywords that would allow us to track online conversations surrounding specific topics, including anti-LGBTQ+ rhetoric, anti-comprehensive sexuality education (CSE) sentiments, concerns about the over-empowerment of the girl child, the perceived softening of the boy child, and anti-feminism narratives.

This query aimed at capturing a wide spectrum of discussions around the themes mentioned above, especially the oppositional stances that often accompany these topics. While we initially sought to incorporate anti-CSE within these categories, we recognised that isolating this theme specifically within the broader discourse of family values would be challenging. As a result, we decided to allow these themes to emerge organically from our broader queries, understanding that the complexity of these issues would be reflected in the data we gathered.

By structuring our query around these targeted subtopics, we sought to create a robust data set that could accurately reflect the ongoing conversations and provide valuable insights into the narratives shaping public discourse in Kenya.

This approach helps in understanding the current landscape and anticipating future trends and challenges in the realm of online discussions of social issues.

The Intricacies of Structuring Queries

The next step involves structuring the query to yield the desired results. This is where the technical aspects of query design come into play. A well-structured query must balance specificity and breadth, ensuring that the data returned is both relevant and manageable.

In our initial attempts at query design, we encountered a common challenge: striking the right balance between too much irrelevant information and insufficient data to draw meaningful conclusions. This challenge highlighted the necessity of Boolean  Operators—a specialized form of search logic that used operators like AND, OR, NOT,  and quotation marks (“”) to refine search results. Boolean Operators were essential in query design because they allowed us to combine keywords strategically, filter out unnecessary information, and hone in on what truly mattered.

For example, in our project, Boolean operators became particularly relevant when  isolating discussions around specific themes such as “family values” and “LGBTQ+  issues.” Our goal was to capture conversations that addressed these sensitive topics without being overwhelmed by unrelated content. By combining terms like “Family AND Values”, we could refine our searches and focus on the nuances most relevant to our research. This logic-based approach sharpened our queries and enabled us to sift through large amounts of data efficiently, finding specific information that directly informed our analysis and strategic goals.

We also expanded the query by providing additional terms and phrases aligned with each aspect of our focus areas. For instance, our search provided a more  comprehensive and accurate picture of the online discourse by incorporating slang like ‘GayZ’—a play on ‘Gen Z’ used to suggest that the generation driving anti-government  protests is primarily made up of LGBT+ people, in an attempt to discredit the protests and sow discord—and other colloquial expressions commonly used in these discussions.

Recognizing Kenya’s unique linguistic landscape, we also translated key terms into  Swahili. We introduced terms specific to the Kenyan context, such as references to Article 45 of the Kenyan Constitution, which addresses family rights. These adjustments helped make our queries more culturally and contextually relevant.

Moreover, we grouped keywords on a matrix to find a balance between those more implicit and could bring too much “noise” and those more explicit. This careful categorization helped us avoid the pitfalls of overly broad or overly narrow queries. For instance, combinations like “Family AND Values” or more specific pairings such as  “Amerix^[1] AND BetterTogether” allowed us to capture targeted conversations without being bogged down by irrelevant data.

Filtering out the “noise”—irrelevant data that could skew our results—required a keen understanding of the online landscape and an iterative process of refining our queries. By systematically identifying and excluding terms that introduced noise, we were able to sharpen our focus on the conversations that genuinely mattered to our research.

Once a query is developed, continuous testing is required to filter out noise and ensure that the data returned is relevant. This process is best carried out collaboratively, as involving diverse perspectives allows for more thorough problem-solving.

Conclusion

Effective query design is both an art and a science. It requires a deep understanding of the subject matter, a mastery of technical tools, and a collaborative spirit. In the context of our project, query design has proven to be a critical component in enhancing our ability to monitor and analyze online conversations in Kenya.

As we refine our queries and expand our research, the insights will undoubtedly contribute to more informed and strategic communication efforts within the Kenyan ecosystem. By mastering the intricacies of query design, we are better equipped to navigate the digital landscape and amplify the voices that matter most in the ongoing dialogue around social issues.

***

^[1] Amerix is the X handle of a popular self-fashioned Masculinity Consultant who often posts on Twitter selling a particular brand of toxic masculinity characterized by strong anti-feminist rhetoric. He usually pushes his content with the hashtags #MasculinitySaturday and #BetterTogether.