Follow Gregory Piatetsky, No. 1 on LinkedIn Top Voices in Data Science & Analytics

KDnuggets Home » News » 2015 » Sep » Opinions, Interviews, Reports » Workforce Data Science: Does Talent Development Increase Performance Over Time? ( 15:n31 )

Workforce Data Science: Does Talent Development Increase Performance Over Time?


Large organizations spend millions on training, coaching, mentoring, re-training and competency development programs. Business ROI with these predictive projects is very significant, here we are sharing some of our findings as they may challenge some concepts we hold so closely.




Our predictive workforce assignments yield staggering results; saving / making businesses millions of real, measurable dollars.  Often this yield is in a single project. Business ROI with these predictive projects is so significant, I wanted to share some of our findings as they may challenge some concepts we hold so closely.

Large organizations spend millions on training, coaching, mentoring, re-training and competency development programs.

We do this believing employee development programs can train someone to better or even great performance. It seems to make sense.  And, the only option a business has after hiring someone is try to develop employees to greatness. Better training, better managers, better raises, better perks, better culture, better benefits, more time off, better work life balance, more and more and more that the organization needs to do, to prop up and hopefully “develop” the employee into being a better performer.  But does it work?

Here’s the thing.

Data Science Studies: Do Development Programs Help Increase Performance Over Time?

We recently completed two analytics studies quantitatively analyzing sales rep. performance. Among other goals, we analyzed sales rep. performance over time – both before and after training is completed.  We documented their sales performance as new hires, during training, and finally after they reached full self-sufficiency in their role.

  • Both sets of analyses were for sales role at vastly different types of companies
  • The first sales group consisted of financial advisors
  • The second sales group sold web / email and internet connectivity into consumer homes
  • Each business measured real sales performance (not performance ratings) over time, during and beyond formal and informal training.
  • Both companies had made a significant, long term financial investment in formal training and coaching development programs, to help sales reps to optimize their sales performance

Results:  Two Data Science Studies

Sales Rep. Performance Does Not Measurably Increase Over Time — Even With Substantial Training and Development by the Business

What we found is perhaps shocking to many, though we see this time and again in our predictive analytics work across many roles.

For the purposes of this paper, we’ll show graphs from the work we did studying Underwriter sales performance.

  • Figure 1 shows one example of Insurance Underwriters. In this case, performance is measured by sales and bookings – so a lovely quantitative measure.  This is not performance review information, this is actual sales performance in the job.
  • Breaking out data by Performance Clusters, we begin to see the true stratification that exists inside this group.  Figure 1 shows the five levels, from the 5% percentile to the 95% percentile performers.
  • Important lessons are evident:
  • Relative Performance Levels are evident very early in tenure, even in the first quarters
  • Growth after initial training is only gradual (if at all) – with the exception of the top performers after quite some time
  • At the individual level, we do not see underwriters crossing categories or increasing after the first year. Performance patterns are set in very early.
  • The first year, or even six months of performance appears to lay the baseline for the rest of an underwriter’s performance

In both of sales projects referenced here, analytics results showed that sales rep performance did not measurably increase over time – despite multi-millions being spent on development efforts including:  training, coaching, competency development and the like.  

talent-analytics-employee-dev-graph1

Figure 1: Performance Level over time for Top 5%, Top 25%, Middle, Low 25%, Bottom 5%

In both of sales projects referenced here, analytics results showed that sales rep performance did not measurably increase over time – despite multi-millions being spent on development efforts including:  training, coaching, competency development and the like.

Top performers began as top performers – and continued to be top performers.

  • Note that the development and experience only made a significant impact on top performers sales and bookings over time.
  • Average performers began and continued to be average
  • Bottom performers began and continued to be bottom.

Training might have had a short term impact, or a small impact.  But long term business performance was “consistent” from the beginning throughout their tenure – with the single exception of the top performers who excelled even more once nurturing was applied to the ideal nature for the role.


Sign Up