Allen Hornung is manager of Talent Analytics & Strategic Workforce Planning at Enbridge
Work in large companies isn't limited to official reporting lines between an employee and their manager, or even collaboration within teams. Modern organizations are complex and social. Executives need to understand how people truly work together to accomplish results. This is the difference between the official structure we know, and the (often more important) informal networks we don't.
By leveraging big data and analytic skills, it is possible to examine these networks. Like an X-ray below the surface of an organizational chart, organizational network analysis (ONA) uncovers the invisible lines of communication and information flow – the lifeblood of most corporations.
Why does it matter that some people communicate across departments and others don't? Do we care that one employee has a colleague network different in size or shape from the others? Once mapped, it's possible to make changes to employee networks that optimize your human capital investment.
If an employee is an information bottleneck, finding that blockage allows workflow redesign to improve productivity. Or if collaboration is needed across departments, ONA ensures the right employee, who holds an oversized share of those connections, is tapped for the role.
Large-scale change initiatives like mergers and acquisitions require identification of champions to distribute messages and feed sentiment back to planners. Through ONA you can find the employees capable of influencing a broad audience (not just those seen as popular by managers).
What if your business depends on innovation and partnership across disparate groups? Gauge the strength of existing connections and adjust project staffing or introduce events intended to build new relationships.
Companies have also leveraged ONA to identify emerging talent or support succession planning. Potential replacements for sudden departures can be selected from those with the same structural profile in the network. Often these people aren't on existing succession lists but could be seamless transitions.
These examples are largely the responsibility of corporate human resources – other cases require broader involvement. For example, an office move prompts a new workspace layout – ONA can help optimize the floor plan. Even in the case of a full reorganization, ONA is of use.
Through ONA, several common network roles are identified. For example, a "broker" provides the rare connection point between two separate departments; a "central connector," the glue within a given team. Some employees are "key influencers," while others are "peripheral" and hold less sway with peers.
The science of mapping and understanding social networks isn't new. Data from e-mail and social networks like Facebook and Twitter are now prevalent and easily accessible. As a result, the past two decades have seen an explosion of activity in network analytics research. However, companies exploring ONA need to decide how to get the raw network connection data that feeds the analysis.
Data scientists analyzing online social networks might exploit the "digital exhaust" already available from electronic communications. Metadata from e-mails, instant messages or corporate social networking sites like Yammer are readily available. Some companies track the flow of e-mails and phone calls within reporting structures, but also throughout the organization to reveal those employees who are well connected across departments. However, organizations quickly run into issues of data privacy and trust. If employees don't expect IT to be tracking this type of information or use it in this way), it could cause reputational damage.
A more accurate and transparent approach is to survey employees, directly asking about their network. The wording of questions depends on the nature of the network being mapped: communication, information flow, advice and mentorship or social connections. Often there will be a combination of intended purposes.
Once raw data has been collected, a variety of tools (both open source and for purchase) are available to analyze and visualize network maps. Several advisory firms also offer solutions supporting ONA from start to finish. The right internal expertise is also needed to get the most from ONA.
Professionals with backgrounds in data mining or analytics – like actuaries – are valuable resources for managing ONA projects. Traditionally thought of as quantifying insurance risk and premiums, actuaries' mathematical modelling abilities are increasingly sought to harness the value of big data. Actuaries with additional training in machine learning and predictive analytics (my personal background) are especially well positioned to support this type of work.
The study of internal networks promises corporate leaders new insight into their workforce. It also offers an opportunity to understand–and optimize–how work really gets done in modern, complex organizations.
Executives, educators and human resources experts contribute to the ongoing Leadership Lab series.