3 Predictions on the Post-ChatGPT World of HR Analytics

Author: Kinsey Li

ChatGPT is becoming the new shiny tech buzzword dominating our social media feeds. This AI-powered chatbot is described to have the superpower to combine swaths of text harvested from the internet with human assistance to execute tasks in natural language. ChatGPT can answer a question, craft a poem, answer interview questions, or write a report for you. 

ChatGPT represents a tipping point for Artificial Intelligence (AI) technology. AI has been around for a while, but ChatGPT made the 'What's in it for me' for AI clear to us. It has done so by transforming years of data accumulated on the internet into intelligent natural language answers to our questions. 

With this in mind, I couldn't help but wonder how HR analytics will evolve in the post-ChatGPT world. Here are my three predictions on how the HR analytics profession might transform. 

Prediction 1: Analytics translation will overtake data integration/modeling as the key skill for HR Analytics professionals

Imagine you can get access to any data you think you need, in a matter of a few seconds. 

A simple question on new hire turnover will no longer take time to analyse, it will only take a question of "Is there any department that has unusually high new hire turnover?" and then potentially followed up with a question of "What did the leavers for department X say about their working conditions?". 

Existing technology solutions are already giving us a glimpse of what's possible, Power BI's Q&A functionality allows you to ask your data questions. Products like ThoughtSpot enable search-driven analytics. As these tools and technologies collectively evolve and morph into autonomic systems, HR analytics' role will fundamentally change. 

In this brave new world where processed people data is readily available and abundant, the skills of asking the right questions and filtering noise from signals will become increasingly critical. As HR analytics professionals, we will still need to understand the nuances of the fields and metrics, but what's more important is the ability to consult and compel. 

The barrier to entry for the HR Analytics profession may also change, data integration and modeling skills will be beneficial, but may no longer be critical for HR analytics positions. We may see more horizontal movements from HR consulting roles into analytics and insights focused positions. 

Day in the life of an HR analytics professional now: 

"Some consulting, some dashboard building, some modeling and a lot of data collection, cleansing and integration"

Day in the life of an HR analytics professional in the future: 

"Hey Analysis Robot, I just come off a call with Production and they are concerned about burnout. Please build me a deck on employee wellbeing by Production departments and highlight high-risk areas by demographics mix"

Prediction 2: HR Analytics professionals will play a bigger role in safeguarding equity and fairness of opportunities

The launch of ChatGPT did not ease any worries related to ethics and biases of AI output. ChatGPT, like many AI-powered products, mirrors human biases as it connects years of biased human input on the internet. 

I couldn't help but then wonder when we are in a position to ask questions about our people data. Will the AI be intelligent enough to identify existing bias or will we need to train the AI to identify its own bias?

If I ask AI the question "Are our promotion numbers balanced across minority groups?", will it be smart enough to understand that 'minority' may not be a singular measure of gender / ethnic group, but instead a collection of statistically significant variables that range from socio-economic backgrounds to lived experiences? Will it be smart enough to be able to find clusters of concerns in a non-biased way?

AI opens up great opportunities for the HR analytics profession as it enables us to combine more, test more, and see more at speed. HR analytics professionals will inevitably face the question - "Are the answers well thought through and free of any institutional biases?"

It will be our role, as HR analytics professionals to train the machine what 'equity' and 'fairness' means. 

DE&I Q&A for an HR analytics professional now: 

Question: "Are our promotion numbers proportional to our population by gender and ethnicity? "

Answer: "There is no significant promotion disparity in promotion rates by genders and ethnic groups."

DE&I Q&A for an HR analytics professional in the future: 

Question: "Hey Analysis Robot, I want to understand if our promotion decisions are biased in any cluster in the organisation"

Answer: "Looks like the high performing AI engineers in Julian and Darryl's departments with similar experiences aren't promoted as fast as the other employees in the same cohort, you can consider what support Julian and Darryl may need to help them."

Prediction 3: Workforce planning will no longer involve monumental effort

Workforce planning is not easy. Particularly for larger organisations where the decisions and accountabilities are dispersed across functions and geographies.

Operational workforce planning for larger organisations typically involve monumental effort in balancing and reconciling finance, operations and HR metrics. A seemingly simple question of "How many people we may need in 3 months with UX Design skills?" will take significant time and effort to model and answer accurately. 

The difficulties in answering questions like this affect our abilities to design effective HR programs such as recruitment sourcing and internal upskills initiatives. 

Technologies like ChatGPT can change the game. These tools have the potential and power of tapping into various data sources at speed and scenario model at scale, while transforming HR and Finance analytics' role in workforce planning from modelling to consuming and assessing. 

A glimpse into the future:

ChatGPT like tool: "You will need 200 extra UX designers in 6 months based on all projects in thepipeline, 50 can be cross-skilled internally from other departments. For the other 150, you can choose to recruit in the market or outsource to external agencies."

HR Analytics professional: "What's the most cost-effective way of acquiring the 150 UX designers?"

ChatGPT like tool: "Based on current market condition and recent cost per hire, we recommend acquiring 60% of the candidate via direct sourcing and the rest via agency contractors. "

This leads us to the bonus prediction - where will HR Analytics teams be in the post Chat-GPT world? 

Bonus Prediction: Blurred lines between functional analytics teams may enable the formation of self-organising teams

As AI-powered tools like ChatGPT continue to connect siloed data points across functions and systems. The ogranisational design for analytics teams may change fundamentally, and the specialisations in different areas (e.g. Finance, HR, Operations, and Marketing) may become more interchangeable. 

It might become easier for analytics professionals to work across a range of different problems, regardless of whether they are aligned to HR, Finance, or marketing. Equally, the proposed solutions from analytics professionals may become more informed as the analysis variables will no longer be limited by function and specialisations.

What can organisations do to get ready for the Post-ChatGPT world?

In my opinion, the three actions that may help organisations get ready for the implementation of autonomic systems are: 

1. Align internal data definitions - aligning data definitions and having a consistent data dictionary established and adopted across all functions

2. Improve efficiency in data collection and storage across functions - minimising duplications in data collection and storage across functions and creating unified data storage guidelines

3. Invest in cross-functional resources to establish consistent ways of working with data - dedicating resources to establish unified ways of working with data across functions and systems

In Gartner's words, in its visionary Hype Cycle for Emerging Technologies research - 

Reference

Gartner_Inc. (n.d.). What's new in the 2022 Gartner Hype Cycle for Emerging Technologies. Gartner. Retrieved February 9, 2023, from https://www.gartner.com/en/articles/what-s-new-in-the-2022-gartner-hype-cycle-for-emerging-technologies 

Chatgpt is a tipping point for AI. Harvard Business Review. (2022, December 14). Retrieved February 9, 2023, from https://hbr.org/2022/12/chatgpt-is-a-tipping-point-for-ai 

Shankland, S. (n.d.). Why the CHATGPT AI chatbot is blowing everyone's mind. CNET. Retrieved February 9, 2023, from https://www.cnet.com/tech/computing/why-the-chatgpt-ai-chatbot-is-blowing-everyones-mind/