I recently read an article published by MIT Sloan Management Review called “Leading with Decision-Driven Data Analytics.” The premise is straightforward, and one I agree with – if you find a purpose for data instead of finding data for a purpose, you will lose almost every time. The authors of the article coined this simple principle “decision-driven data analytics.”
At Personify, we know that merely using the data you have on hand won’t generate analytics that provide meaningful and actionable insights. It often leads companies to focus on the wrong questions. However, decision-driven data analytics define the decision to be made and what data is required to make that decision. Once we identify the correct questions and required data, a data scientist can transform the unknown into the known, identifying appropriate solutions that lead to meaningful improvements in outcomes.
We’ve outlined the differences between data-driven and decision-driven analytics below:
Four critical steps are required to avoid preferences and biases and make the shift to decision-driven analytics:
- Identify multiple alternative courses of action
- Determine what data is needed to evaluate alternative courses of action
- Collect and evaluate the required data
- Select the best course of action
Let’s relate this to talent acquisition. Imagine a large company has collected data over time about its candidate pool and employees. The company is experiencing a higher-than-normal attrition rate. The company looks at its employee data and sees that many self-terminating employees came from companies with unlimited sick days. The company changes its benefits policy to reduce attrition rates to include a new benefit–unlimited sick days. In this example, no alternative actions were considered, and the decision maker’s preferences led to the result. This is an example of using the data on hand to make data-based decisions without asking the right question; Why are employees leaving the organization in the first place?
To answer the question “why are employees leaving,” with decision-driven analytics, the company must first pull together data from different sources to identify potential causes of churn. For example, it might look at exit interview results, performance ratings, succession planning, compensation, benefit offerings, promotion rates, the external market, and social media platforms such as Glassdoor. It might then survey employees as to how important each potential cause is to employees. Then it might analyze all the data to look for patterns, correlations, and sentiment that identify the root cause of the attrition leading to more meaningful mitigation actions.
Turning talent acquisition into a decision-driven platform delivering actionable insights leading to better talent solutions means identify leading and lagging indicators to turnover. At Personify, we help companies combine data such as event-driven growth, a new contract sales funnel, a loss of contract, retention rates by proximity to facility, engagement, performance, and market compensation to get a better sense of expediting the recruitment process. After all, companies can only go so fast without predictive data.
The goal is to recruit the right people who perform at the highest level and stay the longest. And to remain ahead of the curve, companies should recruit ahead of the need. Making decisions without understanding the whole picture causes many companies to focus on the wrong questions while flying blind with data or act under the influence of preexisting beliefs and incentives. In the example above, having a large data set didn’t equate to the best decision. Likewise, making hiring decisions without understanding factors of high performers, what incentivizes churn or retention, and how to apply that data to a repeatable process incorporating performance data can lead to sub-optimal decisions delivering poor results.
To ensure your investment in a data analytics program produces meaningful results, take a decision-driven approach. It will help you focus on the questions that matter most, tie results to action, and eliminate assumptions on how the world works.