Machine Learning Times
Machine Learning Times
EXCLUSIVE HIGHLIGHTS
Effective Machine Learning Needs Leadership — Not AI Hype
 Originally published in BigThink, Feb 12, 2024.  Excerpted from The...
Today’s AI Won’t Radically Transform Society, But It’s Already Reshaping Business
 Originally published in Fast Company, Jan 5, 2024. Eric...
Calculating Customer Potential with Share of Wallet
 No question about it: We, as consumers have our...
A University Curriculum Supplement to Teach a Business Framework for ML Deployment
    In 2023, as a visiting analytics professor...
SHARE THIS:

CONTINUE READING: Access the complete article in LinkedIn, where it was originally published.  

8 years ago
Predictive Analytics World For Workforce Reflections

 

Here are some reflections and key takeaways. Please note this is not exhaustive! I did not see every presentation.  A sign of a great conference is how torn attendees are between concurrent presentations. Good to note that PAWWF had this key indicator. I thought about building a model to help me with that decision. Then I realized that any target variable (e.g., quality) would be analogous to performance reviews and would not have enough variance: Exceeds Expectations! To make up for my shortfalls, I would ask other attendees to comment, fill in the gaps, add to my comments, or just tell me where I am wrong.

From Why to How to Sharing and Developing Innovative Practices.

The field of Workforce/People analytics (WFA) is growing and evolving at a rapid pace. We have clearly moved way beyond the ‘why should we do it’ to the ‘how should we do it.’ The field is in a unique place in its evolution.   There are a diverse number of very talented people working in the field and this will undoubtedly drive innovation. This makes WFA exciting and made the conference exciting!

The Difference Between Predicative and Reporting.

This debate in the field might hurt WFA more than it helps. Maturity models might not be a great way to look at the field or your current position/trajectory. For clarity, if you are having a conversation with executives about the difference between predictive and reporting, you are having the wrong conversation! You should be talking about the issues (i.e., problems and opportunities) facing the business, how to prioritize these based on value, and how your efforts can help with those issues. This will drive what WFA should/is focusing on to add value.

To expand, if you look at the issues facing the business and how to apply WFA, it will undoubtedly need a collection of tools, techniques and processes. This includes a dynamic data strategy that will integrate multiple data sources that can grow with you.  Advanced analytics to create better decision making (or myth busting). And ways to communicate and deploy your findings to drive the business including reporting and other decision tools. A couple of presentations did an amazing job with this. They even showed how their time spend had changed. Essentially, in WFA there is a need to be strategic both about the business, and how to focus time and efforts to create some repeatable solutions throughout our workflow. The strategic and repeatable part was explained with some nice examples.

Tools, Techniques and Vendors

Focusing on the issues will require a collection of tools, techniques, and vendor collaborations. There were some exciting trends provided by a number of speakers including some key notes, panels and plenary sessions. The tools are evolving to help WFA execute techniques and also to help WFA choose which techniques. For example, the ability to do some ‘model trashing’ faster. Exploring multiple techniques is a must. This means one of the biggest advances will be in our ability to scale and support our talented people. This will have a key impact on a number of stakeholders:

  • As an Executive or CHRO it will be important to find a way to help enable your WFA team and to provide your WFA team with the necessary resources, support and tools. There is some great research on the relation between job demands and how this matches resources provided. If you have highly talented people without the resources and support including help and training time, they are likely to leave as someone else will have those tools and resources for them. This is evident in WFA. They are usually asking because they believe it will help them bring value to the organization. In most cases it will make them more effective and valuable to the organization. Having the right resources/tools upfront may seem more expensive because the cost is more visible, but in the long run not having the right resources/tools is always more expensive becasue of hidden costs (e.g., employee time, lost employees, missed value adding/cost reduction opportunities, unnecessary repetition, etc.). Involving the WFA team in the tech budget/spend might be a good idea.
  • As a WFA team, there is a need to be business driven and creative. Business Driven means building your burning platform or sense of urgency as was highlighted.  Essentially, build asks in a business case to help show the value added. This will also help with being more strategic. Being creative means we can’t have or need everything. There is debate about what tools are most appropriate and if WFA needs big data tools and the elephant in the room. Special note, the small few using it seem to like it and are having success with it! Have a look around the organization to see what you might already have. It is usually easier to borrow something you already have.  Other functions are having many of the same data issues and most likely want/have similar tools.
  • As a Vendor, it will be critical to have the end user(s) in mind and understand the HR IT ecosystem and how you fit in. What are the key issues your product will solve, and how can you help solve the other issues that your product does not solve? There were some stellar presentations as usual on work and employee attrition, but a key idea about needing the data out of the system in order to integrate it with other data was mentioned. This issue will continue to grow. Think about it during product development.

How to Proceed Forward 

Prioritization and iterative or agile development keeps showing up. There were great presentations on using iteration and letting the issues drive. Rarely are technology solutions stable, nor do we get it correct on the first try. Building ideas so we can iterate was very evident. Some showed how pilot testing and research design was useful and needed. Involve the business and ask for feedback during the process when it won’t impact the project. It will save you time in the long run. We are going to make mistakes and we all need to learn along the way, the key is to make them fast, cheap and learn quickly.

If there are mistakes or misinterpretations those are all mine!

Derrick McIver, PhD researches, teaches and consults in the area of workforce analytics.

By: Derrick McIver, Assistant Professor, Western Michigan University
Originally published www.linkedin.com

Leave a Reply