Advancements in cognitive technology allow companies to better analyze predictive data, but low-tech behavioral science techniques and culture research are still essential to successful talent acquisition.
IBM’s discoveries over the past 30 years in talent acquisition prove that company culture has a direct impact on engagement, and engagement is a driver of performance. We’ve also found that it is possible to develop a culture profile for teams who have the most success within an organization.
Even with advancements in how cognitive technology derives predictions from HR data, there will never be accurate predictions of candidate success without a profound understanding of behavioral science and organizational culture.
The problem with performance data
When applying cognitive analytics to RPO, the ultimate goal is to create an ideal culture model that would predict candidate success based on company or industry. The current approach involves linking existing data to desired outcomes, based on quantitative and qualitative research.
The biggest hindrance to creating standard models of success is performance data. Cognitive analytics systems need objective and consistent performance data, such as sales or production numbers, to create predictive models. However, much of the performance data that feeds cognitive analytics is subjective. Currently, there’s no way to benchmark subjective performance data.
An example of subjective performance data is the way managers rate their employees. Usually, managers do not rate their employees at the extreme ends of the scale even if the rating is deserved, or they rate employees at the same level regardless of actual variation in their performance.
To minimize the impact of subjective performance data, companies must focus on their talent management practices, grounding them in objective behavioral science. Culture research can show where a company is now and the direction it needs to go, and talent management can move a company — and its employees — in that direction.
Cognitive systems need behavioral science and human input
Cognitive systems such as IBM Watson Recruitment enable larger, more robust collection and analysis of data sets. Recruiters can take more variables into account, enabling the cognitive system to link data from across a company to get a picture of what success looks like for that organization.
However, a cognitive system cannot pick the variables — at least not yet. For now, humans need to steer the iterations of analysis and find objective performance data to feed into cognitive systems.
In order to get the true benefit of a cognitive system when it comes to predicting candidate success, look first to the data you already have.
- Do you have objective performance data to provide, so a cognitive system can learn what great looks like?
- Do you have an objective assessment tool to determine if a candidate is a culture fit?
- Do you have a profound understanding of your own organizational culture, rooted in evidence-based behavioral science?
If you can answer yes to these questions, you’ll get the most benefit out of a cognitive system. But without these critical base tools, all of which require human input at the front end, your cognitive system will be operating with flawed information, leading to flawed outcomes.
To learn more about the link between culture research, traditional behavioral science techniques, cognitive technology and their roles in predicting candidate success, download our 2017 State of the Industry paper today.