1. Introduction
Social Network Analysis (SNA) is the use of networks to understand social structures. Networks are characterized by nodes and edges where nodes are entities and edges are relationships between those entities. SNA has been applied in a large number of fields including sociology, economics, anthropology, marketing, linguistics and computer science to name just a few. SNA is also used commercially, for example to analyze and improve organizational networks and for network weaving in the non-profit space. Here are some concrete examples of SNA in practice:
- A marketing team using SNA to analyze social media data such as from X/Twitter
- A consultant using SNA to understand the internal and external connections at a large hospital and using that data to improve care and outcomes at that hospital
- A regulatory authority using SNA to understand the resilience of a banking system in the face of a potential financial crises
We live in an increasingly interconnected world and networks lie at the heart of so much in this world. Relationship mapping is important in order to understand how things are connected and also to drive change, where appropriate.
Whether you are a research student, a consultant, work in Government or at a non-profit it’s important to have a powerful SNA tool when engaging in an SNA project. This guide outlines what to look for in an SNA tool and also some tools to consider.
2. What to Look for in an SNA Tool
Some of the key features to look for in an SNA tool include:
- Network visualization capabilities: Arguably the core feature of an SNA tool is the ability to visualize networks. It should include one or more network layout algorithms as well as the ability to size, color and filter both nodes and edges. Ideally it also allows you to analyze sub-networks easily by applying one or more filters and saving multiple different views for a single network.
- Metrics: The most commonly used metrics are measures of centrality such as In Degree, Betweenness Centrality and PageRank. You will likely also want some edge metrics such as the ability to differentiate between mutual (i.e. reciprocated edges) and non-mutual edges as well as core network metrics such as Density, Average Shortest Path and Diameter.
- Data integration and import options: How easy is it to get data into and out of the tool? You should ensure that your tool of choice supports common file formats such as GEXF and GraphML so that you can easily interchange data between different tools. Ideally, if it’s a cloud-based tool, it also has an API that allows you to write code (or have an LLM write code!) to automate the import and export of data.
- Ease of use: SNA has its roots in academia and historically SNA tools have emerged out of academia where there may not have been the same attention paid to ease of use and an intuitive user experience. Many modern SNA tools don’t compromise on ease of use - they have an intuitive interface but still make powerful analysis possible.
- Scalability for large datasets: You may not have a large dataset now but it’s likely that you will have one in the future. You should ensure that the SNA tool that you select is capable of scaling up to the size of data that you expect to have. It’s important though to understand that most SNA datasets are actually not that large. Generally, selecting a tool that can handle 10’s of thousands of nodes and 100’s of thousands of edges is sufficient.
- Security: If your data is in any way sensitive you will want to select a tool that is secure by design. Desktop tools can be easier to secure but come with the tradeoff of often being difficult to secure and don’t support sharing network diagrams with others. If you opt for a more modern cloud-based tool you should verify its security controls, including that those controls have been validated by an independent third party (e.g. SOC 2 or ISO27001 certification).
- Sharing: Network visualizations can be a powerful way to present data for decision making and increasingly network analysts need the ability to share saved views and interactive network visualizations with others. If sharing networks with others is something that you anticipate being necessary then you should investigate cloud-based SNA tools that support sharing across multiple accounts.
3. Best SNA Tools for Relationship Mapping
The good news is that there are quite a few social network analysis tools available to help you analyze your data. Below we have summarized some of the advantages and disadvantages of a few of the more popular tools.
Tool 1: Polinode
Polinode was designed as a tool for organizational network analysis, i.e. visualizing and understanding the informal networks, collaboration patterns and information flows within organizations. However, it includes a very general social network analysis tool that includes the ability to upload any type of network data so is used for social network analysis more generally as well as organizational network analysis.
Strengths
- Modern user-friendly interface
- Cloud-based so insights can be shared with others
- Unique and powerful features such as the ability to save multiple views per network and apply layers
Weaknesses
- Not well suited to extremely large networks as it is browser-based
- Does offer a free tier but is a commercial product unlike other open source solutions
Tool 2: Gephi
Gephi is open-source desktop software for the visualization and analysis of networks. It uses Java and is published under the GPL 3 licence. It is one of the most popular tools for network analysis, primarily because it’s free and handles relatively large networks. There are also a number of plugins that extend the functionality of Gephi.
Strengths
- Handles large networks well
- Long history - available since 2007
- Sizeable array of network layout algorithms and metrics available
Weaknesses
- Not web-based software so it’s not possible to share interactive networks with others
- Can be challenging to install and maintain given its dependency on Java
- User interface is not intuitive and requires a steep learning curve
Tool 3: UCINET
UCINET is a software package for the analysis of social network data. It was written by Lin Freeman, Martin Everett and Steve Borgatti and is closely associated with the LINKS Center at the University of Kentucky. Its history dates back to the early 1980s. The latest version of UCINET is UCINET 6 and it comes with the NetDraw network visualization tool, i.e. UCINET is primarily focussed on calculating metrics and performing network analysis whereas NetDraw is more concerned with network visualization.
Strengths
- Extremely comprehensive set of network metrics
- Large academic literature and long history of use in academia
Weaknesses
- Does not scale well to medium or large networks (the practical limit is considerably less than 5,000 nodes especially for more computationally demanding metrics)
- Commercial software though generous discounts are available for students and academic users
- Written in Pascal as opposed to a more modern language
- Emphasis is on academic applications and the learning curve is very steep
Tool 4: NodeXL
NodeXL is an Excel plugin designed for the visualization and analysis of network data. Its primary use case is social listening, i.e. tapping into data from sources such as X (formerly Twitter), Wikipedia, YouTube, Reddit and Flickr and integrating with tools such as Brandwatch, Meltwater, Talkwalker and Tweetbinder. It’s often used to teach network analysis in an academic setting.
Strengths
- Relatively simple for beginners to get started with
- Supports all of the most popular network metrics such as Pagerank and Betweenness centrality
- Good for importing data from social network sites and sources
Weaknesses
- Excel-based so difficult to share interactive networks though does have an online graph gallery
- Doesn’t scale well to truly large networks
- Commercial product though discounts are available for students
Tool 5: Pajek
Pajek is a Windows program specifically designed for the analysis and visualization of large networks. It can scale to up to 1 billion nodes with the only limit being the available memory. The history of Pajek dates back to 1996 and it is used extensively in academia.
Strengths
- Supports very large networks - designed specifically for that purpose
- Documentation available in a large number of different languages
- Free software but not open source
Weaknesses
- Targeted at advanced users - there is a steep learning curve
- Requires installation of desktop software and not possible to share interactive networks
- Outdated user interface not designed for interactive visualization
Tool 6: Communalytic
Communalytic is a computational social science research tool for studying online communities and public discourse on social media and includes the ability to both collect network data from those online communities as well as to visualize it.
Strengths
- Built specifically for text and social media network analysis with integrations with Bluesky, Mastodon, Reddit, Telegram, X (formerly Twitter) and YouTube
- Web based tool that is relatively easy to use
- Includes topic and sentiment analysis
Weaknesses
- More geared to the analysis of text data than network analysis more generally
- Online network analysis lacks polished UI and flexibility of other tools
- Student accounts have relatively low limits and professional accounts are paid
Tool 7: Cytoscape
Cytoscape is a Java desktop application that was originally designed for the visualization and analysis of large scale biological and gene networks. It is capable of importing and analyzing more general network data and also supports a separate javascript library for web visualization.
Strengths
- Open source software licenced under the GPL
- Extensive app store with various “plugins”
Weaknesses
- Primarily a desktop application which makes it difficult to share interactive networks with others though some plugins do attempt to tackle this
- Complex UI with steep learning curve
- Limited support as not a commercial project
Tool 8: ORA (Organizational Risk Analyzer)
ORA is software that is developed and maintained by the team at the CASOS Centre at Carnegie Mellon University. It’s specifically designed for dynamic networks, that is to say networks where the nodes and edges change over time. It contains a large number of metrics for general social network analysis, dynamic network metrics, trail metrics, procedures for grouping nodes, comparing and contrasting networks, groups, and individuals from a dynamic meta-network perspective. There are two versions of ORA - ORA-Lite which is the free version available for use with up to 2,000 nodes and ORA-Pro which is a paid version. ORA is frequently used for military and intelligence applications.
Strengths
- Designed specifically for dynamic networks. Contains many advanced metrics and simulation functionality
- Provides insights into risks around networks
Weaknesses
- Complicated to use with a steep learning curve
- Closed source desktop software that depends on Java so difficult to share interactive networks with others or customize
- Dated user interface
Tool 9: NetworkX
NetworkX is a Python package for conducting general network analysis. It is open source software provided under the 3-clause BSD licence. The strength of NetworkX is that it is part of the Python ecosystem and contains a large number of network algorithms as well as graph generators and other utilities such as for importing and exporting network data. It also provides the ability to visualize network data but its strength lies in the algorithms rather than the visualization.
Strengths
- Large number of metrics and utility functions
- Flexibility of being able to write code for custom metrics and approaches
- Open source software
Weaknesses
- Python library so it’s not possible to interact with the network visualization and the visualization functionality is generally quite limited
- Requires knowledge of Python and programming more generally
- Not web-based so no ability to share networks with others
Tool 10: Linkurious
Linkurious is a commercial software solution that sits on top of a graph database such as neo4j and allows you to visualize and analyze network data from this graph database. It is primarily used by enterprises for applications such as fraud protection and information security.
Strengths
- Handles arbitrarily large graphs by querying the underlying graph database
- Insights can be shared with others, i.e. supports collaboration
- Well designed visual interface for querying and analyzing network data
Weaknesses
- Complicated and expensive to setup - typically only an enterprise solution
- Requires a graph database
- Mainly focussed on solutions such as preventing fraud and monitoring information security
4. Tips for Choosing the Right Tool
There are quite a few options when it comes to selecting a tool for the analysis of network data. When working through the selection decision you should keep in mind to:
- Match the tool to your use case and goals - for example if you are focussed on a use case that one of these tools is specifically designed for (e.g. military simulations and ORA) then you will most likely want to start your testing with that tool
- Understand what your budget is - if you have zero budget then you may want to focus on some of the free or open source tools whereas if you have some budget and support is important to you then some of the more commercial tools will likely be attractive
- How much data do you have? If you have millions of nodes and edges then a solution like PajekXXL or Linkurious may be necessary. Most people don’t actually have truly large networks though.
- Do you want to share data and insights with others? If so, a web-based tool that enables collaboration like Polinode may be the best solution for you.
Finally, please remember that you can and should try more than one tool and you may actually end up using more than one tool - for example, it’s possible to export from Gephi directly into Polinode and also from NodeXL into Polinode.
5. Conclusion
Make a short list of your requirements and then a second short list of the network tools that best satisfy those requirements. If the tools on that list are free tools you can get started with them straight away. For the other tools, most vendors (including Polinode) will provide a free trial so you should explore whether those tools will work well for your specific use case too. The key is to experiment and to choose the right tool or tools for you.
One additional resource that may be helpful when assessing Polinode’s capabilities and whether they fit your needs is this video that provides a detailed walkthrough of our network analysis functionality.
Andrew Pitts
Andrew Pitts is the Founder and CEO of Polinode, a leading provider of organisational network analysis software and solutions for enterprises. Andrew founded Polinode in 2013 and it is now used by large enterprises and consulting partners around the world for a variety of applications including: identifying emerging and/or hidden talent; improving collaboration; finding influencers; succession planning; organisational design; and diversity and inclusion. Prior to founding Polinode, Andrew worked in the Investment Banking Division of Goldman Sachs in both Sydney and New York. He enjoys working at the intersection of technology and HR and is passionate about using technology to help optimise and improve modern organisations. Andrew is a full-stack developer with experience building scalable web applications and also has deep expertise in data analysis and machine learning.
