How to Track Social Influence at a Network Scale

For social technology companies working with governments to manage foreign ideological interference, understanding how sophisticated actors manipulate social networks at scale can provide crucial insight for the development of platform governance policy and user experience features.

Through comprehensive case study analysis of international mis/information campaigns from 2016 to 2022, I identify a handful of strategies used by government organizations to achieve influence at a network scale.

The Scale Challenge in Digital Influence Research

Traditional approaches to analyzing online persuasion focus on individual content pieces—specific posts, memes, or arguments. This methodology, while valuable, misses how influence actually operates in networked environments where thousands of coordinated accounts can shift public opinion through structural interventions rather than compelling individual messages.

My research addresses this gap by applying social network analysis concepts to understand what I call "networked influence"—strategies that aim to intervene on network structure itself to promote particular perspectives. Using systematic case study methodology, I analyzed three distinct Russian information campaigns and one grassroots campaign to identify reproducible patterns of network-level influence.

Four Core Strategies for Network-Level Influence

• Networked Contagion (2016 Election Case): Analysis of Internet Research Agency operations revealed how influence spreads through exposure rather than argument quality. During the 2016 election, 80+ IRA employees posted over 1,000 pieces of content weekly, reaching 20-30 million viewers monthly. The strategy succeeded not through sophisticated arguments—many contained obvious translation errors—but through saturated exposure within targeted communities. This demonstrates how behavioral contagion principles from network science apply to opinion formation: people adopt beliefs when surrounded by others expressing those beliefs, regardless of argument quality.

• Relational Brokerage (2020 Election Case): By 2020, the IRA had adapted to platform countermeasures by creating Peace Data, a U.S.-based news website that hired American freelance journalists to write content promoting Russian strategic interests. This illustrates strategic network brokerage—using intermediary nodes to access otherwise restricted network segments. The American journalists served as credibility brokers, filtering Russian editorial perspectives through authentic U.S. cultural knowledge and linguistic competency.

• Network Partitioning (Ukraine Invasion Case): When Russia invaded Ukraine, the government created legally enforced network partitions by blocking Western social media platforms and criminalizing dissenting discourse about the invasion. This network isolation strategy demonstrates how structural control can shape opinion by limiting access to alternative perspectives, creating artificial consensus through information scarcity rather than persuasion.

• Strategic Community Formation (NAFO Response Case): The emergence of the North Atlantic Fella Organization (NAFO) as a grassroots counter-influence network reveals how homophilic clustering enables coordinated response to information campaigns. NAFO members connected based on shared capacity for meme production and opposition to Russian invasion, creating an organic network that could rapidly identify and respond to pro-Russian content across platforms.

Strategic Applications for Social Technology Platforms

This research offers actionable insights for companies developing platform governance, content moderation, and user experience features:

• Early Warning Systems: Understanding contagion patterns enables platforms to detect coordinated inauthentic behavior before it achieves scale. When multiple new accounts simultaneously promote identical perspectives in targeted communities, this signals potential network manipulation rather than organic opinion formation.

• Broker Detection and Management: Identifying accounts that serve as bridges between different communities can reveal both malicious influence operations (like Peace Data) and organic information flow patterns. Platforms can develop more sophisticated approaches to managing these crucial network positions.

• Partition Resistance: Understanding how network partitions function helps platforms design features that maintain diverse information exposure while respecting user preferences for community curation. This requires balancing echo chamber prevention with user autonomy.

• Community Health Metrics: Analyzing the formation and behavior of communities like NAFO provides insights into how platforms can support beneficial collective action while preventing harassment or coordination that violates platform policies.

Methodology and Research Applications

The comparative case study approach I employed involved analyzing network behavior patterns across multiple time periods and influence campaigns, using both quantitative network metrics and qualitative analysis of strategic evolution. This methodology is replicable for studying other large-scale influence phenomena beyond Russian information operations.

Key methodological insights include:

• Network-level analysis reveals influence strategies invisible through content-focused approaches

• Longitudinal case studies show how influence strategies evolve in response to platform countermeasures

• Cross-platform analysis captures how influence operations adapt to different network architectures and policy environments

Business and Policy Implications

Understanding networked influence has significant implications beyond platform governance:

• Risk Assessment: Organizations can better evaluate how their communications might be manipulated or amplified through network effects, informing crisis communication and brand protection strategies.

• Strategic Communication: The research reveals how authentic grassroots movements (like NAFO) can effectively counter sophisticated influence operations through understanding of network dynamics rather than resource advantages.

• Platform Responsibility: The findings suggest that effective content moderation requires attention to network structure and relationship patterns, not just individual content pieces.

The research demonstrates that influence in networked environments operates through structural manipulation rather than persuasive content alone. For social technology companies, this means that user protection and platform integrity require sophisticated understanding of network dynamics, not just content analysis.

By focusing on how connections form, how influence spreads through exposure, and how communities can be strategically isolated or bridged, this research provides a roadmap for building more resilient social platforms that can support authentic human connection while defending against coordinated manipulation.

The implications extend beyond security concerns to fundamental questions about how social technologies can best serve human communication needs in an environment where network structure itself has become a contested space.