Organizational Network Analysis

Organizational network analysis (ONA) is a method for mapping and measuring the informal relationships between employees: who goes to whom for information, advice, decisions and help. Instead of the reporting lines on an org chart, ONA shows how work actually moves, revealing hidden connectors, overloaded people, silos and employees at the edge of the network.

Why does organizational network analysis matter for enterprise HR?

Organizational network analysis matters because most decisions about people are still made from the org chart, and the org chart only shows who reports to whom. It does not show the engineer every team asks before a release, the operations lead who links two regions, or the new manager whose trusted contacts are all still in their old team.

Collaboration is also under strain. Only 29% of employees report satisfaction with work collaboration, according to Gartner's quarterly surveys of about 18,000 workers presented at its ReimagineHR conference in October 2024, and that share has been falling (Gartner's HR research release). ONA gives HR a way to see where that collaboration breaks down instead of guessing.

For enterprise HR and talent leaders, the payoff shows up in a few places:

• Retention: People drifting to the edge of the network, or tied to an overloaded colleague, are early flight risks.
• Succession: A central connector who leaves takes relationships with them, so ONA feeds succession planning for roles that look junior on paper.
• Change: Informal influencers spread a new process faster than a memo, which makes them useful allies in change management.
• Onboarding: HR can check whether new hires have built working ties after 90 days, rather than relying on how they say it is going.

What are the core components of organizational network analysis?

Organizational network analysis is built from five parts: the people and relationships, the data that captures them, the metrics, the roles those metrics reveal, and the map that pulls it together.

Nodes and ties

Each person is a node, and each working relationship is a tie. Ties can carry a direction (who asks whom), a strength (how often or how much) and a type, such as information, advice, decisions, trust or energy. Most projects map two or three types of tie, because the advice network and the decision network rarely look the same.

Data sources: active and passive ONA

Active ONA uses a short survey in which employees name the people they rely on for specific things, such as expertise or approvals. Passive ONA reads metadata from email, calendars and chat tools to see who interacts with whom, without reading message content. Survey data captures trust and quality of help; passive data is current and needs no survey effort. Many teams combine both.

Network metrics

A handful of measures do most of the work:

• Degree: How many direct ties a person has.
• Betweenness: How often a person sits on the path between two groups that would otherwise be disconnected.
• Closeness: How quickly a person can reach everyone else through the network.
• Density: How many of the possible ties in a group actually exist.

Network roles

Metrics point to roles that the org chart hides: central connectors who many people rely on, brokers who bridge teams or regions, peripheral people with few ties, and energizers whose interactions leave others motivated. The same role can be an asset or a risk; a broker is valuable until they become the only route between two departments.

Network maps

The output is usually a visual map in which clusters, bridges and isolated groups are easy to spot, alongside tables of metrics by person, team or location. Leaders tend to act on the map, and analysts on the numbers.

How do you conduct an organizational network analysis?

An organizational network analysis follows the same seven steps whether it covers one team or a whole company.

1. Define the business question: Pick one decision ONA should inform, such as reducing silos after a reorganization or protecting key people during a merger. Deloitte's 2024 guidance on ONA stresses defining what good collaboration looks like for each team first, because broad collaboration is not always the goal.
2. Set the scope: Choose the population (a function, a region, the whole company) and whether you need each person's view or the full network.
3. Choose the data: Decide between a survey, passive metadata or both, and agree which types of tie you will measure.
4. Agree privacy rules up front: Write down who sees what, at what level of detail, and how long data is kept, then tell employees before collecting anything.
5. Collect and clean the data: Run the survey (aim for a high response rate, since missing people distort the map) or connect the data sources, then match records to your HR system.
6. Map and analyze: Build the network map, calculate the metrics and look for patterns: silos, bottlenecks, overload, isolated newcomers, missing links between teams that should work together.
7. Act and repeat: Turn two or three findings into specific actions, such as a new cross-team forum or a backup for a single-point broker, then rerun the analysis in six to twelve months to see what changed.

How is organizational network analysis different from social network analysis?

Organizational network analysis is the workplace application of social network analysis (SNA). SNA is the broader method for studying relationships in any group, from online communities to disease spread. ONA applies the same methods to employees and work, and adds HR data such as role, tenure and location. It is also often confused with the org chart itself.

ONA also overlaps with HR analytics: HR analytics looks at attributes of individuals, such as turnover risk or performance, while ONA looks at the relationships between them.

What are the best practices for organizational network analysis?

Organizational network analysis works best when employees trust it and leaders act on it. A few practices make the difference:

• Be open about purpose and data: Explain what is collected and why before you start. UK ICO guidance on monitoring workers expects employers to tell staff and to run an impact assessment where risk is high.
• Report patterns, not rankings: Share team-level maps widely and keep person-level views to a small group, because a "least connected" list quickly becomes a layoff rumor.
• Never read centrality as performance: A highly connected person may be a key contributor or simply overloaded with meetings. Pair network data with performance and engagement data before drawing conclusions.
• Repeat the analysis: One map is a snapshot. Running ONA on a regular cycle shows whether actions such as new forums or role changes actually moved the network.
• Protect your connectors: Use the findings to share load away from overloaded people and to document what single-point brokers know, which also supports knowledge management.

Organizational network analysis also feeds hiring. When a map shows that a key broker is about to leave, or that two teams lack anyone with the skills to connect them, the gap becomes a concrete role requirement. ONA can tell you which relationships and skills a team is missing, but not whether a candidate has those skills.

That is where role-based skills assessments help: hiring teams can test for the exact capabilities the network is missing, using tests from the Testlify test library, instead of hoping a resume reflects them. Read more on why skills-based hiring is changing talent acquisition.

Frequently asked questions (FAQs)