Network Analysis Tools & Software

Tool Selection Considerations

The right tool depends on:

Factor	Consideration
Data scale	Hundreds vs millions of nodes
Technical capacity	Staff programming skills
Analysis needs	Basic metrics vs advanced algorithms
Visualization goals	Internal analysis vs public presentation
Integration needs	Connection to existing CRM

---

Software Comparison

NetworkX (Python)

Best for: Large datasets, computational analysis, scripted workflows

Strength	Limitation
Handles massive datasets	Requires Python programming
All standard algorithms	No built-in visualization
Highly scriptable	Steeper learning curve
Free and open source	No GUI
Extensive documentation

Installation:

pip install networkx

Basic Usage:

import networkx as nx

# Create network
G = nx.Graph()

# Add nodes (organizations)
G.add_node("ACLU", type="legal")
G.add_node("MutualAid", type="grassroots")
G.add_node("FaithGroup", type="faith")

# Add edges (relationships)
G.add_edge("ACLU", "MutualAid", weight=3)
G.add_edge("MutualAid", "FaithGroup", weight=5)

# Calculate centrality
betweenness = nx.betweenness_centrality(G)
print(betweenness)

# Export for visualization
nx.write_gexf(G, "coalition_network.gexf")

Gephi

Best for: Interactive visualization, exploration, non-programmers

Strength	Limitation
Excellent visualization	Limited to desktop
Interactive exploration	Slow with very large networks
Calculates metrics via GUI	No scripting/automation
Free and open source	Import/export workflow required
Layout algorithms

Workflow:

1. Prepare data in CSV or import from NetworkX (.gexf)
2. Import nodes and edges
3. Calculate statistics (built-in panels)
4. Apply layout algorithm
5. Style nodes and edges
6. Export visualization

Data Format (CSV):

Nodes file:

Id,Label,Type
1,ACLU,legal
2,MutualAid,grassroots
3,FaithGroup,faith

Edges file:

Source,Target,Weight
1,2,3
2,3,5

igraph

Best for: Statistical analysis, community detection, research pipelines

Strength	Limitation
Very fast computation	Requires R or Python
Advanced algorithms	Less intuitive than Gephi
Statistical integration	Visualization more limited
Available in R, Python, C
Community detection

Python Usage:

import igraph as ig

# Create network
g = ig.Graph()
g.add_vertices(["ACLU", "MutualAid", "FaithGroup"])
g.add_edges([("ACLU", "MutualAid"), ("MutualAid", "FaithGroup")])

# Community detection
communities = g.community_multilevel()
print(communities)

# Calculate metrics
betweenness = g.betweenness()
print(betweenness)

Kumu

Best for: Collaborative mapping, stakeholder engagement, presentations

Strength	Limitation
Web-based, no install	Limited free tier
Collaborative editing	Less analytical depth
Beautiful visualizations	Export limitations
Stakeholder presentations	Proprietary
Easy learning curve

Best Use: Presenting network insights to non-technical stakeholders

---

Tool Comparison Matrix

Feature	NetworkX	Gephi	igraph	Kumu
Learning curve	High	Medium	High	Low
Max network size	Millions	~100K	Millions	~10K
Visualization	Via export	Excellent	Basic	Excellent
Scripting	Yes	No	Yes	No
Collaboration	Via code	File sharing	Via code	Built-in
Cost	Free	Free	Free	Freemium
Community detection	Good	Good	Excellent	Basic

---

Data Collection Tools

Survey Platforms

Platform	Network Features	Cost
Google Forms	Basic roster surveys	Free
Qualtrics	Advanced skip logic	$$
SurveyMonkey	Moderate features	Free-$$
Typeform	User-friendly	Free-$$
Airtable	Database + forms	Free-$$

Roster Survey Template (Google Forms)

COALITION NETWORK SURVEY

Your Organization: [Dropdown of coalition members]

For each organization below, please indicate your relationship:

[Organization A]
□ We exchange policy information
□ We collaborate on programs
□ We share funding opportunities
□ We meet at least monthly
□ We refer clients to each other
□ No direct relationship

Relationship strength: [1-5 scale]

[Repeat for each organization]

Social Media Data Collection

Source	API Access	Data Available
Twitter/X	Researcher API	Followers, mentions, retweets
Facebook	Limited (CrowdTangle)	Page interactions, group membership
LinkedIn	Very limited	Connection data restricted

Ethical note: Social media scraping raises significant privacy concerns. See ethics guide.

---

CRM Integration

Action Network

Network capability: Action Builder for relationship tracking

Action Network → API → NetworkX → Analysis
                  ↓
              Gephi → Visualization

Integration approach:

1. Use Action Builder to track relationships between activists
2. Export via API or SQL mirror
3. Process in NetworkX for centrality calculations
4. Visualize in Gephi or Kumu

Salesforce

Network capability: Custom relational objects

Object	Purpose
Contact	Individual nodes
Account	Organization nodes
Custom Relationship	Edge tracking
Tags	Node attributes

Airtable

Network capability: Linked records as relationships

Table	Purpose
Organizations	Node data
Relationships	Edge data (linked records)
Contacts	Individual nodes
Projects	Joint activities

---

Visualization Best Practices

Avoiding "Hairball" Networks

Problem	Solution
Too many nodes	Filter to significant actors
Too many edges	Show only strong ties
Cluttered layout	Use force-directed algorithms
No meaning visible	Encode data visually

Visual Encoding

Element	Encode Using	Example
Node size	Centrality measure	Larger = more connected
Node color	Category	Legal = blue, grassroots = green
Edge thickness	Relationship strength	Thicker = stronger tie
Edge color	Relationship type	Funding = gold, info = gray
Position	Network structure	Clusters grouped together

Layout Algorithms

Algorithm	Best For	In Gephi
ForceAtlas2	General networks	Yes
Fruchterman-Reingold	Small networks	Yes
Circular	Showing all connections	Yes
Geographic	Location-based networks	Manual

Presentation Tips

Audience	Approach
Coalition board	Simplified, key metrics highlighted, clear legend
Community members	Geographic overlay, plain language labels
Funders	Impact metrics, reach visualization
Internal strategists	Full complexity, interactive dashboard

---

Interactive Dashboards

Options

Platform	Features	Cost
Kumu	Web-based, collaborative	Free-$$
Observable	Code-based, D3.js	Free
Tableau	Business intelligence	$$$
Power BI	Microsoft ecosystem	$$
Custom (D3.js)	Maximum flexibility	Dev time

Dashboard Elements

Element	Purpose
Network map	Visual overview
Metric table	Key statistics
Node search	Find specific actors
Filters	Subset by type/region
Time slider	Show change over time

---

Data Management

Secure Storage

Requirement	Implementation
Encryption at rest	Encrypted database/files
Access control	Role-based permissions
Audit logging	Track who accessed what
Backup	Regular, encrypted backups
Retention limits	Delete when no longer needed

Data Formats

Format	Use Case	Tool Support
CSV	Simple import/export	All
GEXF	Gephi native format	Gephi, NetworkX
GraphML	Standard interchange	Most tools
JSON	Web applications	Custom
Pajek	Legacy format	Specialized

Version Control

Track network data changes:

networks/
├── 2026-01-coalition.gexf
├── 2026-04-coalition.gexf
├── 2026-07-coalition.gexf
└── changelog.md

---

Learning Resources

NetworkX

Resource	Type
Official documentation	Reference
Network Analysis Made Simple	Tutorial
Python for Social Scientists	Book

Gephi

Resource	Type
Gephi Quick Start	Tutorial
Martin Grandjean tutorials	Video
Official wiki	Reference

igraph

Resource	Type
Official documentation	Reference
Statistical Analysis of Network Data	Book
R-bloggers tutorials	Blog

General SNA

Resource	Type
Social Network Analysis (Wasserman & Faust)	Textbook
Networks, Crowds, and Markets	Free online book
Coursera SNA courses	Online course

---

Implementation Workflow

Minimum Viable Setup

1. Collect data via Google Forms roster survey
2. Process in Google Sheets (edge list format)
3. Import to Gephi
4. Calculate basic centrality metrics
5. Visualize with ForceAtlas2 layout
6. Export PNG for presentation

Advanced Setup

1. Collect data via Action Network + API
2. Store in secure database
3. Process with NetworkX scripts
4. Analyze with igraph for community detection
5. Visualize in Gephi (interactive) + Kumu (presentation)
6. Dashboard with Observable or custom D3

---

Tool Checklist

For Getting Started

• [ ] Define network scope (what nodes and edges?)
• [ ] Select survey tool for data collection
• [ ] Choose analysis tool based on team skills
• [ ] Set up secure data storage
• [ ] Plan visualization approach
• [ ] Identify learning resources needed

For Ongoing Analysis

• [ ] Establish data collection schedule
• [ ] Create analysis scripts (if using code)
• [ ] Document workflow for reproducibility
• [ ] Train additional staff on tools
• [ ] Plan for data updates

---

Next Steps

1. Review ethical considerations before collecting data
2. Follow implementation guide for getting started
3. Plan interventions based on analysis