Predictive Validity vs. Construct Validity in a Sales Assessment Test

Ed Marsh | Feb 3, 2023

Tl;dr - Great data enables great management, assuming the data actually represent what we believe it does and that it's accurate. Sales assessment tests provide an interesting case study. Whether you use them as candidate assessments for hiring, or evaluations to improve your current team, you must ensure that you understand the methodology and evaluate whether the data is meaningful for your use case. This dive into predictive validity vs. construct validity should help.

What is a Sales Assessment Test?

A sales assessment test is a tool that companies use to measure sales-specific skills and attributes to inform decisions around interviewing, hiring, training, and assigning future and current sales reps. Sales assessment tests should be configurable to report findings appropriately against job specifics and requirements, including the degree of competition, length of the sales cycle, job title selling to, average transaction size, type of sales role (BDR, Sales Manager, Customer Service, etc), and relative pricing. 

For a sales assessment test to have value, though, it needs to be used properly and must actually measure and report actionable information. That means that a sales assessment test that delivers insights that are predictive of sales job performance (rather than characterizations of personality or behavioral style) and a management team with the skills and toolset to understand and leverage the insights to improve individual and aggregate performance.

Why, When & How to Use a Sales Hiring Screening Tool

A sales assessment test can be used to evaluate an existing team to optimize coaching and training, as well as ensure that people are in the best sales role for their capabilities. For current employees, it can be used at any time and then repeated some months later to measure progress against the baseline.

When a sales assessment test is used to aid sales recruiting and hiring sales reps, it should be used in very specific ways to comply with EEOC requirements around pre-employment screening and to provide maximum benefit.

A sales assessment test used as a sales hiring screening tool must be properly configured/adapted to the role, and used with all applicants (internal and external) at the beginning of their application process. This ensures that a predictive measurement of success in the role is the primary criterion considered in selecting interviewees. Doing so eliminates arbitrary factors and possible unconscious bias (for instance gender and other attributes possibly revealed by names) and also significantly reduces the interviewing burden by eliminating a large number of folks whose resumes may appear interesting but who won't perform.

That's different than the improper use by many companies - screening with resumes, interviewing various people, picking a preferred candidate, and then using a sales hiring test to validate their choice.

Contrasting Predictive Validity vs. Construct Validity

The whole premise of using a sales assessment test hinges on a couple of critical questions. First, does it measure factors that actually cause sales success (vs. generally correlate to it), and, second, is there material predictive validity to the results?

It should be obvious but bears mentioning because many personality and behavioral styles test vendors gloss over the fact, that if a test isn't significantly predictive of actual sales success, then it offers little value as a sales assessment test.

Therefore, in order to select a sales candidate assessment that actually improves the quality of sales reps, you need some understanding of testing terminology to evaluate vendor claims. 

Predictive validity vs. construct validity is a very important distinction to understand. These terms of art have specific meanings. Dave Kurlan, developer of the Objective Management Group test and author of the whitepaper The Modern Science of Salesperson Selection, explains the difference in this article, Top Recommended Personality Assessments for Sales.

The short version is that Construct Validity measures how accurately a test measures what it says it does. So, for instance, a sales personality test could measure personality very accurately (high construct validity) but have zero predictive validity in identifying candidates who will sell. Predictive Validity is a gauge of how accurately a test predicts actual job performance

The latter is what you presumably expect a sales assessment test delivers. Will the test reliably predict not only who can sell, but who WILL sell? Many tests don't. To take it to an extreme, a candidate assessment could argue that showing up for work every day is an attribute of successful salespeople, and with high construct validity, tell you which candidates are likely to show up every day. Clearly, that's a specious predictor of sales success!

If you like to dive deep into the weeds, one surefire way to differentiate between tests is to request a vendor's "Technical Manual." They may not have one, or try to stall you or talk you out of it. Or, they may confidently share theirs. There's important info in both their response and in the manual itself.

Of course, you may still use other tools, like sales personality tests and behavioral styles tests, to consider cultural fit, team composition and more. But they don't measure actual sales skills and aren't predictive of sales performance. Only a highly predictive tool will reduce costly recruiting and hiring mistakes, shorten onboarding and ramp-up, and provide an actionable roadmap for training and coaching.

Why Predictive Validity is Critical for a Sales Assessment Test

Pre-employment screening tests for sales (and evaluations for your current team) are only worthwhile if they provide predictive insights - in other words, identify specific reasons why someone will or won't sell when you're hiring sales reps, and coaching insights to help your current folks sell more, respectively.

The value hinges on the test actually identifying factors that are predictive of sales performance vs. those that are somehow correlated but not reliably predictive.

In short, if you're investing resources in the candidate assessment, and its reliability has a direct impact on the time you squander interviewing unqualified candidates, paying folks who will never perform, and sacrificing opportunity in underperforming territories, then the assessment should merit that investment.

Construct validity may confirm you're getting what you're buying (even if that's not what you think you're getting), while predictive validity provides value.

Choosing the Best Sales Testing Tools to Improve Sales Recruiting and to Optimize Your Sales Team

We've all heard, and likely preached, the data mantra. We manage what we measure.

Embedded in that are the two assumptions that we can accurately measure the right things. Those are dangerous assumptions, and many vendors of sales testing tools sell solutions that purport to do that but don't.

Whether your goal in using a sales assessment is to help in hiring sales reps, or in doing some diagnostic sales testing of your current team, the most important decision you'll have to make is what tool to use.

That's complicated when you look at the full panoply of nicely marketed tools that are sold as sales capability tests.

However, it's dramatically simplified when you understand more about the testing methodology and can accurately interpret testing industry lexicon. When you understand the difference between tools that offer predictive validity vs. those that provide construct validity, the options are substantially narrowed, and the choice becomes easy. 

Intrigued? Check out this free trial.

Predictive Sales Candidate Assessments FREE TRIAL