The Era of Using HR Data as Management Metrics
Sumitomo Metal Mining has introduced an HR system that turns group-wide talent data into management metrics. The goal is to centrally manage employee skills, experience, and evaluation data, leveraging it for business decisions.
As labor shortages intensify, the shift from viewing people as “costs” to “assets” and visualizing their utilization is accelerating. Particularly for large companies with group subsidiaries, optimizing talent allocation and advancing talent management are urgent priorities.
However, the question here is not whether using HR data as management metrics is right or wrong. Rather, it’s about how to handle that data.
From the perspective of “reversible management,” turning HR data into management metrics has a major pitfall: the risk of “fixing” people.
The Moment Data Fixes People
Data accumulated in HR systems appears to be objective and fair decision-making material. A skills map instantly shows who is suited for which task. Experience data quickly reveals the best candidate for a project leader role.
But pause and think. The “aptitude” and “track record” shown by data are based on past evaluations. Deciding future assignments based on past data risks overlooking a person’s growth potential and changes in aptitude due to environmental shifts.
In a mid-sized company I consulted for, the top three sales performers were promoted to lead a new business based on the HR system. However, while they excelled at routine sales, they had no experience building a business from scratch, and all three left within a year.
Data is merely a “snapshot at a certain point in time.” Treating it as an absolute metric fixes people into roles and leads to irreversible decisions.
Three Patterns of Losing Reversibility
When turning HR data into management metrics, three points require special attention:
First is “role fixation.” Judging someone as “suited for accounting” based on data locks their career path into accounting, robbing them of opportunities to explore other possibilities.
Second is “evaluation fixation.” If past high evaluations continue to influence future assignments and promotions, there’s a risk of divergence from current performance. This is the trap of the so-called “halo effect.”
Third is “organizational structure fixation.” Designing an organization based on HR data creates a structure optimized for the current workforce composition. But as the business environment changes, so do the required talent profiles. An organization built on past data loses its adaptability to change.
Designing a “Reversible HR System”
So, how can we ensure reversibility while using HR data as management metrics? The key lies in the perspective of “looking at tasks, not people.”
In Sumitomo Metal Mining’s case, rather than fixing each employee’s data as “their ability,” it’s effective to manage it by linking it to “the tasks they are currently handling.”
For example, instead of fixing Mr. A to the title of “Sales Manager,” manage him on a task basis as “currently overseeing sales operations in the Kanto region.” This makes it easier to transfer Mr. A or change his role, increasing organizational flexibility.
Set Evaluation Periods and Observation Points
When using HR data as management metrics, be sure to set “evaluation periods” and “observation points.”
An evaluation period is the timing for reviewing decisions based on data. Pre-deciding “this assignment will be reviewed in six months” or “this role will be reassessed in one year” prevents fixation.
An observation point is a criterion for checking discrepancies between data and reality. Define signals like “expected results are not being achieved” or “feedback from team members is negative,” and check them regularly.
One of my clients introduced this mechanism. After using the HR system for aptitude diagnosis to select new business leaders, they also set a rule to “reassess aptitude after three months.” As a result, two people were reassigned to different tasks, but the transfers were made with their consent. This avoided a situation where decisions based on data became irreversible.
Using HR Data on the Premise of Failure
One of the basic principles of “reversible management” is “design on the premise of failure.” The same applies to using HR data.
The probability that data-based assignments or promotions are “correct” is never high. Rather, assume they will miss the mark and design how to revert when failure occurs.
Specifically, it’s good to decide on the following three points:
1. A confirmation process when data and reality diverge
2. Criteria and timing for reverting assignments or roles
3. How to handle data after reverting (record it as “failure” or as “learning”)
The third point is especially important. If HR data labels something as “failure,” employees will stop taking risks. Instead, it should be recorded as learning, such as “this assignment didn’t yield expected results, but the person excelled in another task.”
I hope Sumitomo Metal Mining’s system functions not merely as “evaluation automation” but as an “experiment in talent utilization.”
The Right Distance Between Data and People
Turning HR data into management metrics undoubtedly contributes to advanced management. However, over-reliance on data risks undermining human potential and organizational flexibility.
What “reversible management” aims for is the right distance between data and people. Use data as input for decisions, but let humans make the final call. And prepare a path to revert in case the decision misses the mark.
I hope Sumitomo Metal Mining’s initiative becomes a step not just toward HR efficiency but toward enhancing organizational reversibility.
If your company is also considering using HR data, why not start by thinking about a “reversible design”?
Comments