While there is a lot of focus on enabling Machine Learning adoption in Talent Operations — a key pre-requisite for ML model creation is the availability of relevant data and a clear understanding of what metrics need to be tracked. Without the relevant metrics in place, any ML initiative is going to be unnecessarily long-drawn and will lead to a loss of credibility for the Data Science program.
So, what is the way forward?
The Human Capital Reporting (HCR) Team needs to take a multi-pronged approach to developing contextual ML-ready metrics -
1) Assimilate Global Standards: Understand existing global standards in reporting, which will come pre-baked with validation across industries and geographies. The ISO (International Organization for Standardization) enables this through a structured and consultative process (ISO — Stages and resources for standards development).
The output of the standards creation process is a document that can serve as a Least Common Denominator (LCD) to be used across industries and geographies. A good reference document is the ISO Standard for “HRM — Guidelines for Internal and External HCR” — ISO 30414:2018 — Human resource management — Guidelines for internal and external human capital reporting. This standard looks at 11 core areas in HR that need to be tracked effectively as part of any HCR approach.
The fact that this is an LCD means that it may not be comprehensive in its coverage and hence needs to be contextualized to your organization.
2) Contextualization: Depending on the organizational vision and regulatory milieu, the LCD needs to be developed further. For example, ISO 30414 has a Diversity Metric which looks at diversity as a percentage distribution of head count across genders, age, skills, and other demographic parameters. However, in the current highly charged DEI narrative, organizations need additional levers to create a truly inclusive environment. A good example of this is the diversity metrics generated by SaaS products like OrgLens which can assess how effectively, for example, women are integrated into the organization’s Social Network (OrgLens | Inclusion Report)
Contextualization can also come in the form of regulatory guidelines. For example, in Service Firms that are talent-intensive — the tracking of protected classes/groups (Protected group — Wikipedia) is critical for the success of any organizational program that seeks to train and evaluate employees.
3) Material Impact on Operations: An additional factor is the concept of “material impact” which is becoming critical for Financial Reporting (https://bit.ly/3QZr5EE) while reporting to regulatory bodies like the SEC. Orgs need to keep a close track of which HCR metrics are considered “material” by regulatory authorities and enable the tracking of these metrics — in addition to streamlining all operations that can improve the same. There are indications that these reporting guidelines will be kicking in quite soon — https://bit.ly/45Sr0Xs.
4) Feedback loop from Data Science Teams: After activating all the levers mentioned above, organizations still need to ensure that the HCR teams are closely connected to Data Science teams to account for missing / inadequate data — discovered during the ML creation process. This deep integration between the Data Stewards and Data Science teams is critical to quickly turnaround predictive solutions that are timely and relevant.
In a nutshell, Talent Leaders need to have a clear strategy to ensure that the HCR pipeline is equipped and leads effectively to provide the foundational support that is critical to support data-driven decisions.
If you would like to have a chat on enabling Talent Metrics Maturity at your organization, reach out to us at hello@hrness.in
#ai #ml #datascience #peopleanalytics #metrics #iso hrness.ai
