Data-driven HR & People Analytics

 

Historically, data has always been the king. Humanity has always made decisions based on data. Data was central to calculations & consequent decisions. But methodologies and relevance were given more importance than the data collection efforts. We are missing it more or less entirely.

Today, we have data all around - emergent data, unnecessary data, and data that has little to no relevance. This data overflow can easily be experienced. It is due more to unplanned than planned events. There could be a creative disagreement about the statement that most of the data is a byproduct of unnecessary processes. Which processes are responsible for producing relevant data, and who are the people capable of determining which processes need to be adopted? How to decide it?

People are confused about the data & its mindful use. Data has no value in itself until it is draped suitably for a particular occasion.

In the world today, businesses are run altogether differently than before; new products need new production processes; technological interface has changed interactions, giving birth to an altogether new type of data. Reels to reals are flooding data warehouses. Technical advancements are coming up with new procedures to first collect, store, and then make it usable with lightning speed. It is still easier with the ‘mechanical’ parts of human existence.

The challenge is seen when it comes to ‘People’ related data. At first, human interaction has changed from ‘people to people’ to ‘people to technology to people’ – changing the entire perspective of social living. What was easy till now has become difficult to interpret. The reason, the technology has inserted its own data into the line of discussion. References are being made out of ‘Reels’ to the real world. Impression is that the ‘reel data’ is real. Making it more complex is AI-created ‘imaginary’ data, full of ‘physical exhibition’ of emotions, as good as humans.

Continuous scientific revealments & revelations about human brain functioning are another angle. Inferences about human behavior under new reveals about human psychology are making it tougher for people who have chosen to be in HR. Any wrong assumption based on ‘inhuman data’ in the disguise of a prediction may lead to the destruction of an employee's career or may even force decision makers to a penultimate decision, leading later to the closure of a company. Building the notion further, research is not done by ‘World Class R&Ds but by capable people. Facilities only assist it.

In this contextual reference to the world, a new HR is taking birth. HR & its ci-devant types are going to be replaced by absolutely another HR that is equipped, on the one hand, with lots of information about humans & their brain and on the other hand, ‘Artificial Emotions’ to choose from the deceptions. HR skills abound today are the outcome of ‘observational psychology’ and ‘Socially drifted’ labor laws. Technology was missing mostly.  It means HR skills shall be challenged to the level that most of them will get replaced by new ones, full of scientific evidence, technical interpretations, and new social laws.

What HR is going to be!

In this dramatic change in & for HR, adopted methodologies to decipher data shall play a bigger role than before, in my view. In data collection and interpretation, which & how of a methodology for its use shall become more critical for HR success.

HR shall be driven more & more by data, but data shall get its relevance from the kind & type of methodologies adopted for data collection, synthesis, and eventually its consumption.

It is easier said than done. HR has to evolve completely from the inside out. New skills need to be grown. Competencies that have been considered irrelevant so far for HR shall breathe more into using new skills.

It means methodology shall be ahead of data to make data consumable. “Decisions are not made on data but their input-friendly outcomes.”

The framework that may be adopted to make HR data-driven and decision-oriented shall include:

1.      Data Collection: get trustworthy signals, not just more data.

Data collection requires a steady stream of information in line with the procedural needs. The second aspect is that the data so collected should be as unbiased as possible to make it input-friendly for better outputs eventually. Any digress in the data completeness may become a reason for erroneous outcomes. In a nutshell, to achieve consistency, unbiasedness, and complete information, due importance should be given process:

i. Identification of a requisite metric: These metrics may include quality, quantity, timeline, cost, and asset utilization.

ii. Defining the ‘Source of Truth’ for the identified metric: For example, if HR has to continuously remain informed about the headcounts and their different subtypes, HRIS can be a source for it.

iii. Standardization of used taxonomies: As for any subject, HR is full of terminologies that have well-defined meanings, but colloquially, these terminologies and meanings are often jumbled up so badly that even HR people have started using them interchangeably. For example, role, job, and position have very definite meanings but are used as if they meant one and the same thing.

iv. HR instrument suitability &applicability: Every HR procedure has certain distinct business requirements to fulfill. These need to be standardized. Just for example, for talent acquisition procedure, mundane instruments like time-to-fill, quality of hire, sources and their efficacy, and interviews and their usable types; business-centric instruments like competency alignment with company values, purpose, vision, different bands & functional needs shall be more suitable & applicable to ascertain the quality of people intake. Creation of non-business-centric instruments not only produces more irrelevant data, but their consumption by users, primarily business leaders, leads to unproductive use of time, efforts, and money. How many people have joined has less relevance than the quality of new joiners.

The Prajjo’s Competency Framework Solution has given a first-of-its-kind solution to organizations to have meaningful business-centric instruments.

v. Data quality controls: Quality controls include all those parameters that provide true meaning to collected data. These generally are completeness, accuracy, timeliness, consistency, and, in some cases, uniqueness.

vi. Efforts for Bias Reduction: Challenging but now possible with many first-of-their-kind solutions which are being offered on the Prajjo platform. It is the quality of people that describes success. Biases, although a part of human sociology, kill the very foundation of success. These are seen at every level of decision-making for people by people. CV shortlisting, interview, promotion, selection for high-value trainings, succession planning, etc. - sure-shot reasons for a company to go downhill as early as possible.

One of the common examples is assigning responsibility to an incompetent family member in family-run businesses. So is the case among professionals who, out of their fondness or other different reasons, intriguingly give more freedom or decision-making authority to incompetents.

Prajjo offers solutions to handle these challenges through its solutions, like JDOAS.

2.    Data Synthesis: turning raw metrics into Causal Insights

Out of three, this stage is not only critical but is central to the entire data processing framework.

Synthesis provides depth to apparent data. Causal relationships are determined based on available data. For example, a company having only 10% of its people aligned to Vision does mean that most of the people are working for functional gains, having no interest in the success of the organization.

Another aspect of data synthesis is to move ‘from metrics to levers.’ It does mean that it is not merely data collection and deriving certain metrics, but to identify and act on specific business activities, i.e., levers that cause those metrics to change, creating a cause-and-effect relationship to achieve end business results. It's about finding what you can control [levers] to influence what you want to achieve [metrics]. To take the above illustration forward, HR is not meant to higher people, but higher people who are aligned with the Values, Mission, Vision of companies. Functional selections only create cultural challenges.

Steps in data synthesis include:

i. Laying down Metrics Hierarchy: Metrics hierarchy means a flow that outlines ‘input to throughput to output’. Output is critical as its quality decides the input quality for the next process until a desired outcome is achieved out of the entire procedure. The entire procedure in HR technology is called ‘Module’.

From an illustration point of view, if a company has not identified competencies in its Competency Framework for its Values, Vision, Mission, etc, inputs for people selection will always be flawed. This is ‘input’. The next part of the process is an Interview that will never help in selecting the right people. This is ‘Throughput’. Wrong selection will only place incompetent people in organizations. This is ‘Output’. And, if this output is input to next business processes, we can very-well imagine final ‘Outcomes’.

ii. Cohort Analysis: Depending upon the company's needs to know various dimensions of operations, different cohorts or groups can be established and continuously monitored for their competencies, behaviors, and performance.

iii. Confounder Control: Naturally and socially, people do fail to differentiate between identical but inherently different group behavior. ‘Control Groups’, therefore recommended to gauge the difference between the groups. For example, collect data for people who were hired basis a comprehensive Competency Framework as recommended by Prajjo’s Competency Wheel Model [a copyright concept] and others based on ‘Truncated -Functional competency framework’.

iv. Establishing Causal Relationship: In business, the impact of one action in a corner if directly triggering a positive or negative impact in a far corner of a company, it thus means these two variables have a Causal Relationship and not mere association. HR has to bring this in its fold. If the exit of a Superior triggers a mass exit behind him, this means that hiring was faulty and was done based on pure functional requirements. If, by hiring people, the basis CW Model skill availability improves, it does not mean the used framework and decisions are causally related.

v. Identification of Indicators: In data synthesis, it is imperative to know whether indicators used are leading or lagging indicators. Generally, leading indicators should be used.

3. Data Consumption: driving decisions and accountability: At the time of data consumption, if it helps in providing the right insight, leading to the right audience and at the right time, the entire process of data collection and synthesis is carried out methodically.

Another aspect that HR has to keep in mind is that the consumption taste is different for different sets of business leaders. One set does not fit all.

Prajjo, on its extensive HR platform, has added new modules, covering all the aspects of data-driven HR practices, thereby bringing HR closer to Business Leaders than ever before.