Sentiment Analysis Helps Companies Listen to the Voice of the Employee

Voice of the Employee Is a Critical Way for a Company to Gauge Staff Morale

Traditionally, organizations have used large-scale, usually annual employee surveys to gauge the morale of their staffs and to understand their concerns about specific issues such as:

• Benefits and compensation
• Working environment
• Workload
• Quality of management

This is known as the Voice of the Employee (VoE).

Companies have also resorted to pulse surveys, which don’t survey the whole workforce but only a subset. This reduces the workload of processing the results. Pulse surveys also allow asking about specific topics that may have only emerged just recently that affect the workforce.

New Digital Sources Contain a Wealth of Relevant Information regarding VoE

Other digital sources of employee attitude data have also come on the scene in the last decade or so. Third-party companies that conduct anonymized surveys of employees and post the results on the Internet don’t just benefit people who want to better understand the conditions inside a company before seeking or accepting a job there. It also gives the companies in question a large new source of data about their current employees’ attitudes. Companies are increasingly taking advantage of this new data to improve their understanding of VoE.

In the past, a challenge in processing this data – including the data in traditional and pulse surveys – was that it was often in an unstructured format consisting of an employee’s unstructured text comments. Analyzing these is simply more time-consuming and involved than, as in many surveys, presenting employees with pick lists as to their level of satisfaction or dissatisfaction with a particular topic. In addition, it is frequently the free-text fields in a survey that allow the employees to make the most informative comments.

What is needed is a technology that can facilitate the processing of employee sentiments contained in these unstructured comments.

Sentiment Analysis Is a Critical Tool for Understanding Employee Sentiment

Sentiment Analysis automates analyzing and making sense of VoE data. In effect, Sentiment Analysis converts the unstructured data (or free text) of employee comments into structured data, with all the advantages of output format predictability and semantic typing of the data. This makes it easy to aggregate the insights found in large quantities of VoE data and get a truly quantitative view of the issues that are most important to the employees.

How Sentiment Analysis Helps to Capture the Voice of the Employee

Sentiment Analysis identifies the emotional attitudes of people based on the unstructured text they produce. The first generation of Sentiment Analysis products simply looked for emotive keywords such as “wonderful” or “terrible.” However, this approach couldn’t figure out exactly which aspect of an entity was being characterized as such.

The more advanced Sentiment Analysis products employ Machine Learning and AI techniques to analyze the linguistic context and derive a much more fine-grained analysis of the sentiment. This is known as Entity-Based Sentiment Analysis.

How Entity-Based Sentiment Analysis Works

Here’s a typical comment that an employee might make in an employee survey:

“I find the promotion opportunities here to be good, but the compensation is not competitive with the rest of the industry.”

There are two sentiments expressed here, one positive and one negative. The output of a Sentiment Analysis product would look like this:

• • Sentiment: positive
  • Sentiment predicate: good
  • Entity: promotion opportunities

• • Sentiment: negative
  • Sentiment predicate: not competitive
  • Entity: compensation

The two sentiments expressed by the employee have been identified, normalized, and transformed into structured data, which can then be quantified and aggregated over a very large set of such comments.

Companies can not only track sentiment about standard HR topics such as compensation and growth opportunities, but also detect new issues (e.g., complaints about some new hazardous work conditions) and address them before they become a bigger problem. Companies can also gauge and track reactions to corporate policies and possibly adjust those policies or inform future ones.

One popular way of displaying sentiment data is via a dashboard which allows tracking of sentiments and emerging themes over time. This gives a good view of the different aspects of Voice of the Employee.

Summary

Sentiment Analysis is an important technology for improving a company’s understanding of its work force and what the concerns, worries, satisfactions, and dissatisfactions it has. It can also measure the attitude toward specific policies and detect emerging issues. Moreover, it can give insight into deeper issues such as the level of employee engagement in their jobs, which is critical to employee retention and ultimately company’s success.