Sales Efficiency – leveraging predictive analytics in the sales recruiting process!!!

1024 684 Pete Furseth

Predictive Analytics in Sales Recruiting

In a recent blog post, we covered an example of the impact of a small change in sales efficiency to orders and revenue. We also listed several methods of onboarding for you to consider for investment. In this blog, we will cover how to leverage predictive analytics to complement your sales recruiting process. This predictive approach will improve your new-hire sales efficiency and will reduce your sales turnover.

Biggest Challenge in Sales Recruiting

One of the biggest challenges in sales is the recruiting of new salespeople. As noted in the prior Sales Efficiency blogs, it requires up to three new salespeople to replace the loss in first-year orders and revenue from the departure of an experienced salesperson. Also, in many industries, the ramp time to a salesperson’s full efficiency is between 15 – 18 months. These factors increase the need to get your recruiting decisions correct. The cost of a bad recruiting decision is high.

We recognize that there are a large number of sales recruiting and human resource assessment companies. Each of these companies has defined its own assessments, which typically focus on 10 – 20 critical attributes for evaluating salespeople. Your candidates’ results are compared to a database of thousands of people in similar sales positions to try to identify good candidates. This can be a good process, however, no matter how many salespeople are in the database, they are not your salespeople with your company nuances and differences.

6 Steps for Your Business

At ORM we recommend a complement to this assessment process. Here are the six key steps to tailoring it to your business:

  1. Calculate the sales efficiency of each salesperson to provide a baseline for analysis.
  2. Define which salespeople are high performers and designate them as such. This designation can be based on overall performance and can also be weighted with factors such as accelerated ramp rates or tenure. The goal is to define the individuals who you would like the new hires to perform most like.
  3. Have 100% of your existing team take an assessment test which will provide the foundation for the predictive analytics assessment. If your team is very large, a statistically valid subset of your employees would also be acceptable.
  4. Analyze the data from the completed assessments correlating the responses to the high performance vs. non-high performance designation. There are several techniques that can be used to evaluate the data. In the example outlined below, we tested several and then selected the Classification and Regression Tree approach to assess the critical attributes.
  5. Evaluate the proposed candidates with the High-Performance Assessment Tree to ensure the proposed hires meet the decision tree criteria for hire. This will provide an independent second opinion based on your own high-performance salespeoples’ attributes.
  6. Remember that this is a validation process and it should be used to complement your hiring process. In the event that there are differences, we recommend you review in light of this data and then rely on your experience and recruiting judgment.


The following is an example we performed at ORM which evaluated 82 employees of which 35 were designated as high performers and 47 were designated as moderate to low performers. All 82 were given a standardized assessment that measured 17 key attributes. From this data, we used a Classification and Regression Tree approach and we were able to create the following decision tree that defines the key attributes in the Hire or No Hire decision. The data is interpreted as follows:

  • There are 17 measured attributes, however using only 4 key attribute scores will allow you to make a Hire, No Hire decision. Each of these areas has a scoring range of 0 – 100 points and in this case, if the candidate’s Intensity score is lower than 30 you get a No Hire recommendation. Next are the Self-Protection attribute, if the candidate has a score of less than 39 the recommendation is No Hire, even if the candidate’s intensity score was greater than 30. The Assertive attribute is next and if the candidate’s score is above 53 you get a Hire recommendation. However, if the candidate’s score is less than 53, but has an Interpersonal Trust score over 80, he or she still gets a Hire recommendation.
  • When the following decision tree is applied to the existing 82 employees, it accurately predicts performance 95% of the time. When applied to the 82 employees this decision tree results in 78 of the 82 salespeople being correctly designated (95%) as High Performers or Moderate to Low performers. 33 out of 35 performers were correctly designated with 2 moderate or low performers being incorrectly classified.
  • Now that we have an accurate assessment model we have a framework to evaluate future candidates. We can have all prospective candidates take the same assessment and then use the decision tree to predict their performance. This prediction can then be used as an additional input to the hiring recommendation.

High performance assessment tree

What to Conclude

In conclusion, recruiting salespeople is a very expensive and critical activity for sales management. Leveraging predictive analytics to complement your existing hiring practices will provide you with an objective assessment of your candidates and increase the number of high performers you recruit. This will, in turn, improve your sales efficiency and reduce your future turnover. All of which will make achieving your order and revenue goals more manageable.