Descriptive, diagnostic, predictive, and prescriptive analytics are fairly common terms and you’ll see this breakdown in many different discussions of analytics. But, I want to go a step further and add another one – I can do that because this is my guide :). In the new world of people analytics, it will also be critical to gain skills in what I am going to call ‘cognitive analytics.’

Cognitive analytics is still a bit of a loose term and not necessarily an official one, but I am going to use it here to generically group together any analytical approaches that go beyond traditional data analysis and either attempt to mimic human-like intelligence or that do not require explicit programming of formulas or relationships to be followed in the calculations. Cognitive analytics may include concepts like artificial intelligence (AI), machine learning (ML), and natural language processing (NLP) to analyze large amounts of structured and unstructured data (numbers, text, images, audio, etc.) and look for insights and patterns that might be missed by more traditional methods.

This field is already transforming the way we do people analytics. With these new approaches and techniques we can tap in to sources of data that have gone ignored and make advanced analytics more accessible to more people. People analytics traditionally relied on structured data (e.g., employee surveys, quantitative performance reviews), but we have so much information already available in the form of employee sentiment, resumes, job descriptions, performance reviews, exit interviews, learning and development progress; all of which was going mostly ignored because there was just too much, it was too unstructured, and the traditional methods weren’t able to catch the nuance patterns and relationships in that data. Cognitive analytic approaches however, if applied appropriately and with an understanding for their (severe) limitations, can help us do more than before in people analytics. We can learn from employee sentiment how the organization is doing. We can identify skill gaps and needs without laborious assessment processes just by leveraging the already available jobs, learning, and performance management data. We can get a better understanding of engagement, retention, and well-being to provide a more holistic and whole-human based understanding of people – rather than trying to summarize people with only basic demographics or unreliable survey results. We can identify talent to bring into organizations in ways that look beyond resumes that can biased by the limited opportunities made available to people in the past and instead look at unique experiences, transferable skills, and diversity of thought. There are a lot of exciting possibilities yet to come for those in a people analytics career who embrace the advanced analytics skills needed to welcome cognitive analytics.

Advanced People Analytics Skills

Having a foundational grasp of core AI concepts like machine learning (ML), deep learning, and natural language processing (NLP) will provide a broader context for cognitive analytics and help you choose the right approaches for you. You will need to learn about different areas depending on the type of data you want to leverage (e.g., NLP is best if you are going to leverage a lot of text, but computer vision will be best if you are going to analyze a lot of photos or video). These are some advanced people analytics skills you could consider building if you wanted to embrace cognitive people analytics:

• Natural Language Processing (NLP) Fundamentals: Learn the basics of NLP techniques that can help you understand the meaning of text data (e.g., word frequencies, sentiment analysis, topic modeling). There are online courses and tutorials that introduce you to the fundamentals of NLP and how it can be used to analyze text data.
• Machine Learning (ML): Grasping core concepts like supervised and unsupervised learning, classification, and regression algorithms is critical for building the models that power cognitive analytics.
• Deep Learning Techniques: While not essential for every project, familiarity with deep learning architectures like neural networks will expose you to cutting-edge techniques used for complex tasks like image and text analysis.
• Computer Vision: Learn about computer vision techniques for extracting insights from images and videos. Techniques include image recognition and classification which allow you to automatically identify and categorize objects within images, or object detection and tracking which let’s you detect and track the movement of objects within videos.
  • Take caution: These approaches do have some very real possibilities for people analytics applications. For example, object detection could be used to conduct employee movement pattern analyses in a factory to identify and remove safety risks, but they should be used with caution. Most employees will likely find it creepy to have their movements tracked or to have you cataloguing photos of them.
• AI Ethics and Bias: AI is becoming more powerful and widely used. But remember, those of us in the people analytics field are talking about real people and very real, life impacting work decisions. Understanding ethical considerations and potential biases within algorithms is crucial for responsible development and deployment of cognitive analytics solutions. Don’t do any type of cognitive analytics without an ethics and bias assessment first (we’ll talk more about the importance of ethics in a later chapter).