An org chart shows roles, departments, and reporting lines. Day-to-day work rarely fits that diagram so neatly. People may go to someone other than their manager for advice. A decision may depend on one specialist. Two neighboring departments may barely share knowledge.
Organizational network analysis (ONA) examines those working relationships and the informal networks they form. It complements the formal structure with evidence about how people interact at work.
What is organizational network analysis (ONA)
Organizational Network Analysis is a method for studying connections among people and teams within an organization.
Organizational network mapping represents employees as nodes and their interactions as ties. The analysis considers the number and direction of those ties, how persistent they are, what they concern, and where each person sits in the network.
Organizational network researcher Rob Cross defines ONA as a way to measure and visualize patterns of collaboration within and across organizations. Those patterns can show information flow in organizations, including how decisions, expertise, and influence move beyond the formal hierarchy.
ONA does not replace the org chart. Formal roles and working relationships answer different questions.
Why an org chart is not enough
The formal structure describes assigned roles and reporting lines. The network shows how help, knowledge, and decisions move among people.
One person may manage a department while the team regularly turns to someone else for advice. Responsibility for a project may be spread across several roles, yet key decisions still pass through one employee. Two departments may sit next to each other on the chart but share very little knowledge.
Microsoft also describes the org chart as an important but simplified view of an organization. Actual communities and working patterns may not follow formal divisions by department, location, or management level.
Standard reporting may miss risks such as:
- critical expertise sits with one person;
- one employee is handling requests from several teams;
- departments have no reliable way to transfer knowledge;
- decisions repeatedly pass through a single bottleneck;
- a team that looks strong on paper is isolated;
- an important informal leader lacks a matching role and support.
What data is used for ONA
Surveys
Employees are asked whom they approach for help, advice, information, or approval. Surveys can examine specific types of relationships, but the results depend on the questions, participants' memory, and their willingness to answer.
Work communication metadata
Some systems analyze anonymized data about meetings, email, and messages. This can show the frequency and distribution of interactions, but the existence of a communication does not explain its meaning or value.
Work actions and signals
Work actions provide another source of data: employees thank colleagues, suggest ideas, report problems, ask for help, or confirm expertise.
No source provides a complete picture. Every conclusion needs context, transparent collection rules, and careful interpretation.
What ONA can reveal
Informal leaders
When colleagues regularly turn to someone for knowledge, support, or help with difficult questions, that person may be acting as an informal leader regardless of their title. This is an observed interaction pattern, not a measure of employee performance.
People who connect teams
Some departments would have little contact without the people who connect them. Knowledge and agreements pass through these employees, so their position in the network matters during growth, reorganizations, and cross-functional projects.
Overload and dependence on key employees
A large number of ties says nothing on its own about performance. It may reflect influence, overload, or the way a process is designed. If too many requests, approvals, and decisions pass through one person, several processes may slow down when that person is unavailable.
Isolated teams and weak ties
The network shows which groups interact often and which operate largely on their own. That pattern may point to an organizational barrier or an information gap, but it has to be checked against the team's function.
Expertise used in day-to-day work
A job title does not always capture every skill an employee has. Requests for help, ideas, and reported work problems can show what knowledge colleagues seek from someone during the observed period. This is a signal of demonstrated or sought-after expertise, not a complete assessment of the person's capabilities.
Recurring problems
When problem signals are collected consistently, recurring issues can be traced across teams and processes. These overlaps support a hypothesis; they do not confirm it without further investigation.
How HiveHR collects organizational signals
In HiveHR's organizational network analysis software, organizational signals come from actions employees take within the platform.
In the basic workflow, employees can:
- thank a colleague for specific help or a contribution;
- suggest an idea;
- report a work problem.
Over time, those actions form an observed network:
- An employee takes a work-related action in HiveHR.
- HiveHR records the tie and the context of the action.
- Related actions begin to show recurring ties and themes.
- A manager gets data that can help frame questions and hypotheses about how work gets done.
Unlike a long one-off survey, this data accumulates through employees' ordinary use of the platform. The network updates as new signals appear.
The number of thank-yous is not a measure of an employee's contribution or performance. Interpretation must account for the content of the actions, the employee's network position, the nature of the ties, and the team's working context.
What can become visible in 30 days
During the first 30 days of a pilot, an initial interaction map may begin to form if enough actions have accumulated to distinguish recurring ties and themes. The map can be used to examine:
- who regularly helps colleagues and transfers knowledge;
- whom people approach most often for expertise;
- who connects different departments or groups;
- whether a process depends on one person;
- where overload may be building;
- which teams have few connections;
- which ideas and problems recur;
- which capabilities appear in observed work actions.
Thirty days of data is not an organizational diagnosis. It is a starting point for observation, helping a team frame questions and choose hypotheses for further investigation.
From there, change over time matters more than a single snapshot. Teams can examine how the network shifts after a reorganization, a new project, a change in manager, team growth, or a new process. Microsoft, for example, uses ONA to analyze changes in collaboration after reorganizations, the move to hybrid work, and other major organizational changes.
How to use organizational network analysis responsibly
Network metrics are not employee performance metrics.
Someone with few ties may be doing important independent work. High centrality can reflect influence, overload, or a poorly designed process. An isolated team may have a problem, or its position may simply match its function.
ONA results should therefore be:
- considered alongside the person's role and work context;
- discussed with managers and teams;
- used to test hypotheses, not to automate employment decisions;
- examined over time rather than as a single snapshot;
- collected under rules that employees can understand;
- handled in ways that protect personal data and restrict access to the analysis.
ONA should help explain how work moves through a team and where support may be needed. It should not be used to find the employee with the highest score.
Frequently asked questions
How is ONA different from an employee engagement survey?
An engagement survey captures employees' opinions and perceptions at a particular point in time. ONA examines the structure of relationships and interactions. The two methods answer different questions and can complement each other.
How is ONA different from People Analytics?
People Analytics is the broader field of analyzing data about employees and the workforce. It may cover hiring, retention, performance, learning, compensation, and other areas. ONA in HR is one method within People Analytics, focused on relationships, interactions, and flows within an organization.
Is ONA a form of employee surveillance?
That depends on how the data is collected and used. Responsible ONA requires a clear purpose, transparent rules, restricted access to results, and protection of personal data. HiveHR bases its analysis on actions employees take within the platform, not on hidden monitoring of private communications.
Can ONA be used to assess employee performance?
Network data is not a performance rating. It can add context to a manager's view, but a person's network position always depends on their role, tasks, team structure, and the observation period.
How much data is needed for analysis?
There is no universal number. It depends on team size, the intensity of interactions, and the type of signals. A pilot needs enough actions to reveal recurring ties and themes; conclusions should not be drawn from isolated events.
See the real structure of your team
HiveHR collects organizational signals from thank-yous, ideas, and reports of work problems. Recurring ties and themes can be used to test hypotheses about knowledge transfer, overload, and collaboration across teams.
Run a 30-day pilot with a single team to establish a starting point for observation and decide which ties and hypotheses need further analysis.
Start a pilot