The best leaders test their opinions with data
Leaders everywhere have one thing in common: they all know what personality traits drive success in their organizations ─ or so they think. There is an overabundance of opinions on this topic, especially in the sales arena, where often the person (even more than the product) is the true differentiator for the customer. But don’t blindly accept my opinion, use Google! A quick search of “top sales personality traits” returns over 1.9 million results with most including ambition, confidence, competitiveness, sociability, etc., as the drivers of sales success. More specifically, a well-cited Harvard Business Review research article found that ambition, assertiveness, and pessimism were 3 of the top 4 personality traits differentiating high-performing sales people from their low-performing co-workers.
While this is definitely a great place to start the conversation, and these drivers may even be accurate in some scenarios, upon closer examination you will find many of the prominent search results are based on opinions of current/former leaders, opinions of existing sales professionals, or at best some aggregation of archival industry data. These beliefs are based on observable behavior. For example, a leader may notice that Bill in sales is really ambitious and extraverted; however, until we compare his personality to his actual sales performance, we will not know how much each trait is contributing to or limiting his success.
This is a topic we have extensively researched and ultimately built a successful product around, but that doesn’t reduce the need to educate the masses. We recently engaged with a CEO of a large sales organization to test this theory. He was unique in that he built the company from the ground up, is still the best sales guy, and not only has an in-depth knowledge of the market history, but a clear and defined strategy regarding how the company is evolving. He had a full understanding of what creates success ─ from the salespeople to the associates throughout the entire organization.
To conduct our study, we allowed the CEO to determine which personality traits were important in his organization by discussing a key sales position with a group of subject matter experts (SMEs) and asking them to 1) independently rate how much of each trait (e.g., ambition) is needed, and 2) how much weight each trait should be given, relative to all others, when trying to drive high performance in the role. We then created a secondary model by using actual employee personality and sales data. This process is accomplished by analyzing incumbents’ assessment results in conjunction with high-priority performance metrics to objectively determine what characteristics are driving success. Lastly, we compared the two models ─ letting the “data speak” to views of the knowledgeable SMEs in the room. This comparison led to a rich discussion about the true nature of the job and high performance.
The following graphic contains some of the results. The top three characteristic levels and weights identified by the SMEs are shown with green arrows (i.e., opinion-based model) and the top three characteristics and weights identified by employee personality and sales data (i.e., data-driven model) are shown with blue arrows.
There were a few notable takeaways from this exercise.
First, there was very low consistency of characteristic importance between the two models. In fact, for the statistic nerds out there, these models only exhibited a correlation of r = .12 (n = 1,912) with a 100% difference in the top eight characteristics. The opinion-based model seemed to parallel the Google results with the inclusion of ambition and competitive fierceness in the top 3; whereas the data-driven model seemed to take a less traditional route with self-reliance comprising almost 10% of the model and moderate (not high) sociability being the third-most important characteristic. This tells us that what we think is important and what is actually important can be two extremely different things.
Second, employee-model fit was analyzed. More specifically, in today’s economy, applicants can come at a premium and therefore models are needed that predict performance as well as allow for high applicant flow. The data-driven model allowed for over 30% of employees to be recommended for hire whereas the opinion-based model resulted in only 6% of employees being recommended. This shortage of candidates can be a major constraint for any organization, but in this specific scenario, it’s an indicator that the specific combination of personality characteristics desired by the SMEs may not be present in the applicant pool.
Finally, we returned at a later date to examine the potential impact of these models on the organization. Those recommended for hire by the data-driven model exhibited a 24% increase in sales and a 34% increase in retention. In contrast, those recommended for hire by the opinion-based model exhibited a 25% decrease in sales and a 30% decrease in retention. These positive results resonated with those in the room, as their approach could have resulted in a significantly less profitable organization.
This leader is now armed with information that will undoubtedly support his organization’s continued growth. I would challenge all leaders to think critically about the behaviors they are selecting/training on and ensure they have a strong theoretical and empirical relationship with their organizational KPIs. As illustrated in this real-world example, the reliance on market-level survey research, or even your own opinion, can be counterproductive to success. When possible, use your ideas as a starting point and examine your own data to ensure your initiatives will result in the desired outcomes. Anything less may result in your organization being led in the wrong direction.
For more information on how Infor uses behavioral assessment data to make more predictive hiring decisions, please review our white paper on the methodology of performance profiles.
Bill Gerber, Director, HCM Data Science and Analytics, Infor
- Talent Science
- North America