Talent analytics is a relatively new and rapidly evolving function at most organizations, many of which are still figuring out where their small but growing analytics teams fit in with their broader HR function. Our ongoing research at CEB (now Gartner) finds that while the vast majority of organizations are investing in talent analytics, very few are seeing results yet.
In David Creelman’s opinion, voiced in a blog post at TLNT, one mistake many organizations are making in developing their analytics teams is appointing a data scientist to lead them. While a talent analytics leader needs an appreciation of data science, a team headed by a specialist runs the risk of focusing on the most interesting data problems rather than those with the greatest business impact:
Here’s what can go wrong if you hire someone mainly for their data analysis skills:
- They’ll lead you down the path to making unnecessarily large investments in data and technology infrastructure — and postpone producing actionable results until that is in place.
- They’ll overlook simple solutions in favor of complex ones.
- They’ll spend months on interesting analysis that does not have a prominent business impact. …
The cart in people analytics are the tools and techniques of analysis; the horse is the business issue. If you hire someone whose main expertise is in the tools, you are putting the cart before the horse. The leader of people analytics is a business person who is given the mandate and the time to look into addressing people issues by delving into the available evidence. They need an appreciation of the tools to do this, but their concern about business impact and general problem solving skills matter more than their expertise with big data.
Another important point to consider here is that the analytics functions at most organizations are still immature, and don’t have enough high-quality data to benefit from the complex models and analysis a data scientist could provide. In our latest survey of talent analytics executives, 70 percent cite data quality as the biggest barrier to doing their work, so their key challenge now lies in building the right relationships throughout the organization to collect better data. Without first improving data quality and credibility, applying advanced data science is wasteful at best, and at worst could lead to erroneous conclusions.
Finally, searching for a data scientist to head the analytics function can be an exercise in futility simply because demand for these experts already vastly outstrips supply. Many employers are looking for PhDs with years of experience to fill data science jobs, but in fact, few such dream employees are on the market, and organizations often don’t need candidates with such extensive qualifications. This labor market reality is motivating organizations to look for new ways to build their data capabilities: Airbnb, for instance, has taken an innovative approach to addressing the data skills gap by launching its own “data university” to develop these skills internally.