The indispensability of Soft Data in HR Analytics / People Analytics


Analytics is considered to be very much a desktop centred job. Those who love working with data, experience the pleasure of drowning in its chaos, while mining for eureka moments using statistical softwares and coding. Thanks to big data – there is a lot of data out there, for such excercises in insight mining. This data is also being continuously compounded almost exponentially every day.

But, HR Analytics is different because it tries to make sense of people and how they can be managed within an organization. However many quantifiable things you may know about a person, including their age, height, salary or sales targets achieved per quarter, you don’t really know the person, if you haven’t met and interacted with the person. Desktop centred analyses can often fall short of organizational needs when it comes to the Human Resource Function. This is where soft data becomes important. While it is not possible to meet everyone within a said workforce to corroborate and build upon the existing knowledge about the said workforce, an HR Analytics practitioner, by default, cannot depend only on hard data. “The best Data Scientists Get Out and Talk to People”, according to an HBR article that came out earlier this year.

What is the difference between hard data and soft data?

Lets try and understand it through the example of recruitment and hiring. According to the cambridge dictionary; Hard Data refers to information such as numbers or facts that can be proved. Within the function of recruitment, this data could refer to a potential candidate’s scores, grades, expected CTC etc. As per the same dictionary, Soft Data refers to information about things that are difficult to measure such as people’s opinions or feelings. Within the function of recruitment this data could refer to the candidate’s values and attitudes towards work, time, co-workers, goal setting etc. The decision of whom to hire, depends both on soft and hard data. To get reliable soft data, interviews are conducted, and more often, probationary periods are set, before a contract for employment is finalised.

However, the human resource function is not just about recruitment. With the growing size and complexitiy of workforces, the HR department in organizations of all sizes is now responsible for employee engagement, employee satisfaction, appraisals and performance management among several other things. One thing in common with all these functions of HR is that, they all involve evaluating candidates/employees or people. Research shows that the five components of emotional intelligence: self-awareness, self-regulation, motivation, empathy, and social skills, are twice as important for excellent performance as pure intellect and expertise. These Emotional Quotient components can only be measured through soft data collection.

An expert in HR Analytics will do at least one of two things while trying to generate implementable insights for success in any of the above mentioned HR functions. S/he will either develop new instruments that can capture and quantify soft data like opinions, perceptions and learnings OR s/he will interact with people and employ intuition based on past learnings to gather insights or soft data. A really competent data scientist will employ both these methods on a continuous basis, improving upon both, based on feedback and past implementations.

Psychometric assessments and Job Fit Assessments are examples of instruments that data scientists within the HR fraternity have developed to measure that, which cannot ordinarily be measured for e.g. Job Fit. These tools can be successful in the long run, only if, they are built on top of algorithms that learn and improve continuously. Machine learning can and has made this possible to some extent. However an HR Analytics practioner cannot get away from conducting some primary field research of their own. This can be through conversations, observations, qualitative surveys and interviews etc. This has to be coupled with an intrinsic desire to learn about people and how they work i.e. psychology. However both hard and soft data are equally important.

A good example of how both hard and soft data can work together to implement a successful strategy can be found in the case of the highly successful and unconventional retail chain: Zara. Though not specifically an example from the HR domain, it has a lot of relevance as Zara is also very people/customer responsive. A study, conducted in the early 2000s on Zara’s ability to rapidly respond and fulfill customer wants, discovered that they used PDAs (Personal Digital Assistants) to collect and communicate hard data like orders and sales trends and soft data like customer reactions and the “buzz” around a new style directly to the headquarters. This direct flow of data and Zara’s flat organizational structure ensured that this data was analysed as soon as it was received and converted into products that flew off the racks because they were always on point when it came to the current style trends.

Because HR Analytics is still at the development stage, it is the best time to develop soft data integrations when it comes to the data being collected to formulate HR strategies for the long term. To do this both employees and candidates have to be measured objectively during formal and informal exchanges that they have within and with the organization. These measures have to be put into place during the HR Analytics Adoption stage and be fine tuned through the early deployment stage. To ignore the importance of soft data within HR Analytics / People Analytics is to deploy a strategy or an instrument that will fail to produce the results it is built for.

To learn more about how you can successfully deploy a well rounded HR Analytics / People Analytics strategy within your organization, write to

Related Blogs

Leave us a comment

You must be logged in to post a comment.