People Analytics, Hypothesis Testing, and OODA.
People analytics leverages HR data to assist in decision-making for HR operations. I am convinced that this initiative will transform HR operations in domestic companies going forward.
While analytics is merely a means to an end, as domestic HR systems and strategies undergo rapid change, the measures taken to realize those strategies are unprecedented. Having no precedent means that we must repeat trial and error starting from hypotheses, and in such cases, verifying hypotheses based on data (facts) will be extremely important.
In other words, it can be said that people analytics is a tool for introducing a hypothesis-testing work style into the HR field.
In this article, I will consider people analytics from the perspective of hypothesis testing.
People Analytics as Hypothesis Testing
In an article I posted recently, I organized perspectives for clarifying data analysis themes using a framework called TIHAM. This can also be considered a perspective for organizing hypothesis testing in business, and people analytics is truly a typical example of hypothesis testing.
People analytics often begins with discovering problems in HR operations. This "problem" arises from the gap between the ideal state (To Be) and the current state (As Is). If the ideal state—that is, the state of the organization that HR aims for—is achieved, then no problem exists, but if the ideal state has not yet been reached, there is a gap (= problem) there.
Once it is understood that a problem exists, we explore the causes and effects of that problem, construct hypotheses for solving it, and then actually implement measures. After implementing the measures, we evaluate whether they are producing the intended effects. This process continues until the problem is solved. This movement can be represented in the following diagram.
As a simple example, let's take the improvement of employee engagement.
Employee engagement is an indicator that represents an employee's level of trust in their company and has been attracting attention in the HR field in recent years. There are various reasons for this background, but I believe it is not unrelated to the increasing fluidity of the labor market and the progress in introducing job-based HR systems.
Now, assuming that improving employee engagement is in line with HR strategy, let's assume we find that it is not reaching its target. Since it is not reaching the target, it needs to be improved, but it does not seem to be a problem that can be solved by simply installing a convenient tool.
This is because such problems are rooted in company-specific circumstances such as organizational culture, management style, and job content. Therefore, it is necessary to examine the actual state of the organization while considering hypothetical solutions and improving through trial and error. It can truly be called a hypothesis-testing initiative.
So far, I have organized the concept of the hypothesis-testing cycle. Since I have numbered them from 1 to 4, you might think that it starts from 1. Of course, in many cases, it starts with 1. Problem Discovery, but depending on the situation, I think it is possible to enter from the middle.
Hypothesis Testing is Only Possible with HR Strategy
The ideal state in the HR field involves securing and developing human resources based on talent strategy, or the introduction and establishment of new systems such as job-based systems. These correspond to the Target in the Data Analysis Framework TIHAM framework, and all must be in line with HR strategy. Furthermore, HR strategy follows business strategy.
Therefore, the hypothesis-testing cycle shown above is considered to function only when there is an HR vision and HR strategy.
If you do not know the direction you should be aiming for, you cannot even find the gap between the ideal state and the current state. In this case, you will not be able to find the problem that needs to be solved. Aimless hypothesis testing might yield results when trying out completely new ideas, but it is not clear whether that will become an initiative that has a positive impact on management.
Introduction of the OODA Loop in HR
Having summarized things up to this point, I suddenly thought of something.
Isn't people analytics essentially about incorporating the OODA loop into human resources?
I then pulled out "OODA LOOP, Chet Richards" which had been sitting unread on my bookshelf, and as I read it, I became increasingly convinced of the deep connection between the two.
The OODA loop was proposed by Colonel John Boyd of the United States Air Force and was intended for decision-making situations faced by pilots of fighter jets and similar aircraft in combat.
John Boyd is said to have been a highly skilled pilot, and the OODA LOOP is a distillation of his thought processes and philosophy. This concept is also gaining attention in the business world.
As shown in the diagram above, the OODA loop consists of four processes: (1) Observe, (2) Orient, (3) Decide, and (4) Act. (The diagram is from Irasutoya.)
Looking at this, you can see that it is similar to the hypothesis testing cycle I presented in this article.
The characteristic of the OODA loop lies in executing operations with flexibility while increasing agility and speed to achieve objectives. In combat, while following operational policies, each individual is required to take flexible measures in response to constantly changing surrounding circumstances.
Isn't this the same for HR managers and HR personnel who implement measures daily based on major HR strategies?
In the 'Observe' part of the OODA loop, it is important not just to look around, but to gather the information necessary for situational judgment by any means possible. In the field of HR, this corresponds to grasping the situation within the organization.
As an organization grows, it would be impossible for HR personnel to interview every employee. Therefore, the advantage of utilizing people analytics in grasping the current situation becomes apparent. On the other hand, there are things that cannot be understood through quantitative data alone, so I believe it is also important to combine qualitative methods. Limiting the information necessary for decision-making solely to people analytics would lead to making the means an end in itself.
Organizational Culture Transformation
It is said that fostering an organizational culture is necessary to incorporate the OODA loop into business. "OODA LOOP, Chet Richards" argues that it is important for each individual to have mutual trust and leadership while the organization shows focus and direction. A culture that allows for trial and error based on hypotheses is also necessary.
Are these not the same for people analytics?
Over the past few years, I have been busy introducing people analytics, and I have felt that people are what matter most, and I have felt the importance of cultural transformation. Looking at it from the opposite perspective, by seriously engaging with people analytics and facing issues such as TIHAM head-on, I believe it is possible to change the way we work. This is because incorporating hypothesis testing and OODA thinking into HR leads to breaking away from a style of following precedents.
I would like to continue to realize cultural transformation in this new field.