Predictive Models, Cultural Fit, and the “Good Hire”: 3 Questions to Help You Get It Right

After an extensive search, Big McLarge Inc. hired Regina for a critical marketing manager role with major growth potential. Her interviews were stellar, her experience deep, her potential obvious. This hire was a no brainer.

But after 7 months of limited progress, strong resistance to change, and a cemented leadership culture more impressed by empty phrases then hard truths, Regina found herself wondering about other opportunities. She bolted within the year.

Was Regina a good hire?

Models to the Rescue… Maybe

By the common metric of tenure, Regina was clearly not a good hire. Even if she was not a serial job hopper, a good predictive model might’ve still flagged her as a flight risk early in the process and saved everyone a lot of trouble and money.

Score one for the modelers, right?

Not quite.

Models use past events and current data to predict future states. To make use of models then, we need to understand how the present maps to the past. If the present and past differ in critical, unknown ways, run for the hills.

One glaring example is the 2007-2008 financial crisis in which, among failures, modelers greatly underestimated the likelihood of a dramatic downturn in the housing market (for more on this rich topic, just Google “fat tails” and Nasim Taleb).

But the challenge of properly linking the past to the present also rears its head in the form of “cultural fit”.

Cultural Fit

“Cultural fit” is a slippery term, but it boils down to how well the candidate and the company match on goals, values, and practices. To see how this issue plays out in human capital, let’s return to Regina, our briefly tenured marketing manager.

According to Regina, Big McLarge Inc. hired her to leverage her unique blend of project management and technical skills. By her account, they were seeking cultural change but then blocked those changes. When she grew frustrated, someone else swooped in, and Big McLarge will pay the price.

Regina’s team members also think the company blew it. She gave her direct reports refreshingly relevant feedback and told senior management what she really thought. She embodied what leaders claim they want their culture to be, but she was still cut off at every turn. She upset the applecart so she had to go.

Regina’s hiring manager sees it differently. Sure, she had some genuinely challenging ideas and the motivation to develop them, but she was a bit brusque and didn’t know how to package things. She should’ve stuck it out and put more effort into blending in and finding common ground. She shouldn’t have left but he probably shouldn’t have hired her.

Modeling: The Challenge of Change

Defining a “good hire”, it seems, requires the lens of cultural fit.

And this, perhaps surprisingly, brings us right back to modeling.

If Big McLarge wanted the future to mirror the past, as their ultimate actions revealed, they would’ve been helped by a predictive model using previous history to flag Regina as a “bad hire”.

If instead Big McLarge was truly seeking to hire “cultural change agents”, Regina would’ve probably been a great hire. Strictly following a model so dependent on standard historical data would be a mistake.

Why?

Companies undergoing radical cultural shifts simply don’t yet have historical examples of “the good hires” within that new company culture, the kinds of people who stick around, perform well, and get promoted. These are the very examples that predictive models need in order to learn who will and who will not be a “good hire.”

Three Questions To Help You Get It Right

Predictive models can dramatically improve hiring and talent management decisions. But behaviors and development always take place within some culture, a culture that shapes processes, motivates performance, and drives outcomes. Changes in those cultures also change those processes, motivators, and drivers.

Before blindly applying a predictive hiring tool, remember Regina, consider your culture, and take a moment to answer the following:

Question 1

How has your organization changed in the last three years? Considerations include culture, market shifts, corporate strategy, regional shifts/ relocation, product offerings, talent management practices, and compensation strategies.

Question 2

Have your hiring and talent development practices changed in response to these changes? Considerations include recruiting, screening, interview structure and questions, pre-hire assessments, development plans, feedback processes, job rotation, and mentorship practices.

Question 3

Given your answers to #1 and #2, would the kinds of candidates that have historically thrived before still be a good “cultural fit” today? Who would still fit? Who might not?

Final Thoughts

All cultures change and evolve over time. And some things like technical skills and motivation do predict hiring success across many different cultures and companies. Cultural changes DO NOT therefore automatically invalidate all predictive hiring models.

But models are just one kind of tool to help us make better people decisions. We will get the most out our predictive analytics if we not only learn from the past, as our models must, but also understand how this past relates to our present and future states.

Like this post?

Get our FREE Turnover Mini Course!

You’ll get 5 insight-rich daily lessons delivered right to your inbox.

In this series you’ll discover:

• How to calculate this critical HR metric
• How turnover can actually be a GOOD thing for your organization
• How to develop your own LEADING INDICATORS
• Other insightful workforce metrics to use today

There’s a bunch more too. All free. All digestible. Right to your inbox.

Yes! Sign Me Up!

Comments or Questions?

Add your comments OR just send me an email: john@hranalytics101.com

I would be happy to answer them!

photo credit: <a href=”http://www.flickr.com/photos/84489055@N00/24282576016″>Trencaclosques de Lego</a> via <a href=”http://photopin.com”>photopin</a> <a href=”https://creativecommons.org/licenses/by-sa/2.0/”>(license)</a>