Since the 1980s senior management has been using passive ONA (organizational network analysis) to better understand their company’s health. By analyzing who is communicating with whom, and the frequency and timing of these communications, network graphs can be constructed and analyzed to provide actionable intelligence. These results can be used to provide insights into how effectively different groups are collaborating, who are an organization’s key influencers, and even insights into the effectiveness of company diversity, equity and inclusion (DEI) initiatives. Furthermore, by analyzing changes in communication graphs over time, trends can be detected, the impact of policies ascertained, and even the risk of potential employee attrition determined.
There are two types of ONA – active and passive. Active ONA comes from data that is collected via surveys submitted by employees. Passive ONA is derived from data that already exists found in emails, chats, and other digital communications. Because passive data is always available, it can be used to continuously measure employee sentiments and behaviors. Contrast that with active ONA where data can only be gathered when a survey questionnaire is sent out (which happens once a year or less frequently at most companies).
Passive ONA provides a powerful tool for understanding corporate health, but it only takes communications metadata into account. The content of the communications is not considered during the analysis. And content matters! For example, consider Beth in engineering sending an email to Jill in quality assurance. From a purely ONA perspective, this represents a link between Beth and Jill. However, consider two potential email threads:
“Hi Jill, just wanted to thank you for all your help on getting the release out last night! Great to have another timely, bug-free release deployed”
“Hi Jill, yikes! What a cluster *&! that release was. Yet again. Why do we even bother, when product management keeps imposing unrealistic schedules…?”
Either of these emails can be exchanged between exactly the same two people. However, the two threads seem to indicate very different company environments. In the first, the employees seem happy, and excited about their string of high-quality releases. In the second, they are demoralized and blaming other groups for their misery. Unfortunately, ONA alone cannot distinguish between these two environments, significantly curtailing senior managements’ understanding of their company! However, by analyzing the emotions and sentiments expressed in these communications we obtain a much deeper understanding of company health than we would from using communications metadata alone.
At RSquared, we take passive ONA results as a starting point, and deploy sophisticated deep natural language processing (NLP) models to understand the emotions being expressed by the users in their communications and even the topics (does a thread represent employees chatting about football or critical project scope requirements) being discussed. Augmenting traditional ONA analysis with information about the tone and topic of the exchanges provides a true understanding of how a company is currently functioning. Furthermore, RSquared aggregates and anonymizes the communications to ensure that the company’s privacy policies are observed, but the overall mood of the company can be understood, monitored and improved over time.
In many respects, these metrics can be even more instructive than surveys, providing insights into how all employees actually feel when conducting their daily work, versus an incomplete point-in-time picture of potentially stylized responses.
Be sure to take a look at the brief demo video of the RSquared platform, or contact us at firstname.lastname@example.org, if you are interested in exploring how to get a deeper understanding of your company’s sentiment, engagement, collaboration, inclusion and culture.