Understand who matters in the network, and why.
Influence Audits use Social Network Analysis to examine how stakeholders connect. The result is a clearer picture of structural influence, coordination and dependency, translated into practical engagement priorities.
Each answer below comes from an established measure, computed on your evidence rather than assumed.
Who holds influence?
Degree and closenessWhat the chart shows. A03 has the most direct connections (4). Eleven actors have 3, A01 has 2 and A14 has 1. A07 reaches everyone else in the fewest steps (closeness 0.43).
What it means. Size shows direct connections. Actors with many ties, and actors who reach everyone else in the fewest steps, are structurally prominent, whatever their title says.
How it is measured
Degree counts an actor's direct connections. In this example it ranges from 1 to 4. A high degree shows reach, not authority: titles, resources and willingness to act need separate evidence. Closeness is the number of other actors divided by the total steps needed to reach them all, from 0 to 1.
Who connects separate groups?
Betweenness centralityWhat the chart shows. A07 and A08 hold the strongest bridging positions (0.55 and 0.54). Next is A03 at 0.28.
What it means. These actors are the bridges between communities that would otherwise stay apart. If one of them steps back, information and coordination between the groups slow down or stop.
How it is measured
Betweenness centrality is the share of shortest paths between other actors that pass through an actor, from 0 to 1. The scores here are computed from the 20 fictional links.
Where are the bottlenecks?
Bridges and cut pointsWhat the chart shows. Removing the A07–A08 link would split the network into two parts of 7 actors each. A13–A14 is also a bridge: it is A14's only link.
What it means. Coordination here depends heavily on one relationship. A single link holding two groups together is a single point of failure: strengthen it, or build a second route.
How it is measured
A bridge is a link whose removal disconnects the network; a cut point is an actor whose removal does the same. Here the cut points are A07, A08 and A13.
Which communities are forming?
Modularity and community detectionWhat the chart shows. Two groups of 7 actors each. 19 of the 20 links stay inside a group; 1 crosses between them.
What it means. Groups cluster around shared relationships. A high modularity means each group mostly talks to itself, the hidden echo chambers, and the few links between them carry everything that crosses.
How it is measured
The groups are defined in the example data. The modularity of this split is 0.45, computed from the 20 links. Values above about 0.3 are commonly read as clear community structure.
Who is being left out?
Low degree and closenessWhat the chart shows. A14 has a single connection, to A13. A01 has two. The farthest from the rest is A14 (closeness 0.22).
What it means. Notice disconnected or peripheral actors who may need intentional engagement. In a real audit a peripheral position can also mean the records are incomplete, so it is checked before it is acted on.
How it is measured
Low-degree actors, read against how the rest of the network is connected. In a real audit, a peripheral position can also mean the records are incomplete.
Where should engagement begin?
Fuzzy cognitive mapping and structural analysisWhat the chart shows. Begin by validating the A07–A08 link, then confirming A14's single link to A13, then checking whether A03's four connections are active.
What it means. The patterns above become a focused engagement plan: exactly which key influencers to engage, where and why, so decision-makers know where to focus their attention.
How it is decided
In a real audit, the measures above feed a fuzzy cognitive map built with your team in Mental Modeler, and a structural analysis of the factors that drive the system. Priorities are validated with your team before any is set.
Network measures describe observed or defined connections; they do not, on their own, prove power, intention or causation.
Measured, not assumed.
An Influence Audit is computed, not presented. We apply internationally established methods from network science and economics to your evidence, with economic factors and research criteria set before any data is read. The analysis is ours; you receive the findings.
- CentralityWho holds structural weight in the network: degree, betweenness and closeness centrality.Social network analysis
- StrengthHow much each relationship carries: tie strength, in-degree and out-degree, weighted by frequency and dependency.Weighted network analysis
- GravityHow strongly two actors should interact given their economic weight and their distance, and where real ties fall short of it.Gravity model, economics
- CommunitiesWhich groups form, how separate they are, and which links hold them together.Modularity and community detection
- Key driversWhich factors drive the system and which depend on it, read from how strongly each one influences the others.Structural analysis
- Cause and effectHow factors push and pull on each other, and what changes if one moves, with expert judgement where evidence is thin.Fuzzy cognitive mapping
Networks are mapped and measured in Gephi; fuzzy cognitive maps are built in Mental Modeler.
Understand the network behind your challenge.
Tell us what is difficult to see in your stakeholder environment. We will explore the question, the information you have and whether an Influence Audit is the right approach.