How to Use Behavioral Analytics
for Employee Retention and More


Retaining key talent is getting increasingly complex. Traditional behavior analysis has proven useful to many organizations in finding, hiring, and even retaining quality talent. It also tends to be a time-consuming effort, laborious, and hard to maintain on a continuous basis. However, we have good news for executives, HR teams, and engineering managers to keep pace, and even gain ground in reducing turnover, but also improve team and organizational performance. 

First, we’ll look at what behavioral assessments are, the different levels, and a variety of important ways they can be used. After that, we’ll focus on behavioral assessments and metrics specifically for retention. Combining these can pave the way for the kind of unified effort to where no one in their right mind would want to leave their teams or your organization. A Moonshot? Perhaps, but that’s what 10x’ing is about. 

Our proposition makes use of an AI-driven Digital Assistant to provide data-driven insights “on-demand” to help key team influencers in your organization 10x your retention, operational efficiency, and growth plans.

What is a Workplace Behavioral Assessment?

Workplace behavioral assessments measure an individual’s performance to characterize their trends in areas of expertise, work habits, motivation, and teamwork. They provide executives, HR teams, and managers data-driven insights to guide a variety of personnel and project-based decisions for optimal results.

Hiring managers and HR teams often use tests and interviews to guide their hiring decisions. These only provide snapshots. But, they’re still quite helpful for validating a candidate’s expertise for matching skill requirements. They also provide a good indication of how well they’ll fit in with your team.

Workplace assessments can be used in a lot of different ways to cultivate continuous “win-win-win” outcomes and relationships between the organization, team, and individual contributor. Traditional assessments have used simulations, tests, interviews, and observed workplace behavior.

More and more companies have started using performance analytics to provide a continuous stream of data. There’s some pushback as to the specific value of any given metric, particularly in that all work tasks are not equal. The value of performance metrics increases relative to the variety and amount of time the metrics are tracked. Variety adds context, time establishes trends — and averages to offset outliers for a sort of normalized view.

What are Behavioral Assessments Used For?

This is sort of a loaded question. Behavioral analytics apply in interesting ways at different levels of an organization to win-win-win outcomes. The organization wins, the team wins, the individual contributor wins. Heck, we should add client and end-user wins, too. This isn’t BS or spin, at least not if you’re serious about helping everyone do and be at their best.

This isn’t to say that sometimes situations will arise where someone loses – “Joe, sorry, we need you to put in some overtime to meet the client’s deadline” or “Mary, we can’t authorize vacations for your team this month due to the importance of this upcoming release.” It does mean that win-lose scenarios should decline in frequency.

Let’s look at specific benefits for each of the main roles in which behavioral data analysis can be put to good use – for C-Levels and Investors, HR Teams, and Managers. These can overlap just as maintaining a transparent program open to individual contributors can further amplify its effectiveness.

Three Levels of Value for Workplace Behavioral Assessments

Executives & Investors Human Resources Project & Team Management
Track organization and project performance
Periodic pay, performance, and promotion evaluations
Track individual and team performance
Alignment of organizational objectives and capabilities
Evaluate quality of recruiting venues
Improve resource allocation
Growth planning for each funding stage
Identify individuals at risk of leaving
Recognize developer burnout
Identify organizational skill weaknesses relative to mission and objectives
Customize training programs
Optimize code review pairing and mentoring

Objective data for one-on-one meetings and Objective and Key Results (OKRs)

Some Behavioral Assessment Use Cases:

Executives and investors should know if their capabilities are improving or not so they can funnel the resources needed to help overcome any challenge. As we covered in Digital Assistants for Software Engineering Managers, the cost for an “average” in-house team of 7 developers runs over $1 million a year – with about a third of that being lost to inefficiency.

Scaling up early-stage startups to meet the demands associated with winning Series A, B, C growth expectations means adding teams, expanded budgets, increased inefficiency, as well as balancing core personnel, in-house employees, and outsourcing options. An engineering manager perfectly suited for early-stage and Series A development may encounter challenges scaling up for your Series B.

Investors have a role to play in that they can introduce your engineering manager to other managers in their portfolio who have already navigated specific challenges.

Human Resources, recruiters, training managers, and even software engineering managers may have overlapping roles. Many startups don’t even have an HR manager. Behavioral assessments can help qualify the quality of the candidates you hire by recruiting venue. Quality, in this context, extends to both technical expertise and retention rates, which could prompt comparing:

  • In-house employees vs. Contractors vs. Augmented Teams, etc.
  • Employee referrals vs. Job fairs vs. Headhunters vs. Job boards (by Job Board)
  • University vs. Technical School vs. Boot Camp vs. DIY Learners
  • By Country, like with the SkillValue Report ranking average developer’s technical proficiencies by country, noting that both skills and wages vary considerably.

Project and Team Managers should be no stranger to performance analytics, even if they are still finding new ways to apply them to their teams. This is a good chance to emphasize that behavioral assessments are useful for all levels of management (HR, C-Levels, and even Investors) for use in one-on-one meetings and OKRs.

The real goal of an engineering manager is to help their developers be the best developers they can be – as relates to their project and company vision/mission. One-on-ones provide a feedback loop wherein the team should also be helping their manager be the best manager they can be. That’s a powerful combination.

Behavioral Analytics helps managers to identify what’s challenging each member of their team the most and other fast-fix, high-impact opportunities. Adding OKR’s to the mix, it’s the engineering manager’s role to help their developers 10x their performance. Historically, if managers had access to performance metrics at all, this ends up being a manual, time-consuming effort.

The Great Resignation

More and more people are quitting their jobs. While some say, “It’s been coming” – the quit rate has been trending up since 2014 and gaining momentum. While the Great Resignation’s impact is highest for those in the retail and hospitality sectors, we can see it in the Information sector, too. If you’re curious about any hiring rate, turnover, quit rate, and similar statistics, by industry, just check out the US Bureau of Labor Management’s nifty tool. The following chart shows the official average monthly quit rate by year for the Information sector. For 2021, this only reflects data through August – reaching a high of 2.9% quit rate across all industries.

It’s logical that with all of the uncertainty regarding Covid in 2020 that fewer people would quit.
It could be that 2021’s was just showing a correction, except that the trend is continuing in 2022. There’s a Global Shortage and growing demand for software developers and really anyone with high-end IT skills. You know it’s bad when even China and its 1.4 billion people have a skilled labor shortageto the tune of 5.5 million? No… 55 million.

Now. let’s add companies asking everyone to return to the office after a year’s worth of working from home.

What? You want me to return to the office? Apple’s employees didn’t seem to like that idea very much. Their response probably gave Amazon the drift to let its corporate staff work from home, indefinitely. Have we reached a point of no return? Whatever the case, the idea of finding someone with the exact skills you need, local enough, and willing to work at your office is getting a lot more challenging.

Then, whether they work at your office or remotely, there’s the matter of actually retaining them, improving overall team efficiency, and preventing burnout. We could probably add a few more challenges to the mix, but this should keep us busy.

How Do You Measure Retention?

For simplicity’s sake, retention and turnover are arguably two sides of the same coin. At the very least, I’m sure that the vast majority of us will agree that we want to increase retention and reduce turnover. These days, even that is not so clear, but hey… I’ve noticed that my own personal performance across all imaginable metrics improves by 27% when I don’t watch the news.
Retention rate is arguably easier to calculate than turnover, as there are fewer different states of being “present and accounted for.” Turnover gets into involuntary vs. voluntary, resignations, retirements, avoidable and unavoidable, etc.

The main one to go after, and arguably the one that’s hardest to plan for, are those resignations – team members quitting particularly to take another job. Up to 50% of millennials say they would accept an offer out of the blue that offers better pay or benefits.

The traditional turnover formula is simple enough:

Turnover Rate = # of Separations / Avg. # of Employees x 100

The definition of employee does add a bit of complexity due to the Gig Economy and freelancers, outsourcing, augmented teams, dedicated contractors, and all of the other variations. Not everyone’s an in-house employee these days. And we’re likely to see a lot of companies redefining their hiring policies to better fit an increasingly distributed work environment. That’s coming as a consequence of the Great Resignation.

Instead of basing turnover on employee counts, for long-term projects managers should factor in any long-term individual contributors they rely upon that are not otherwise based on a defined term or scope. Turnover on augmented teams and dedicated contractors who quit will probably have the same negative impact on your project as an in-house employee.

Retention-Risk Indicators and Metrics to Watch

We identified a number of Reasons, Trigger Events, and Observable Indicators in, Why Software Developers Leave? It’s helpful to distinguish between the occasional bad hair day/s and what are likely to be more pronounced trends and changes in behavior indicating one of your team members is about to leave.

The top reasons why people leave a company include a lack of advancement opportunities, pay and benefits, poor team fit, management issues, and inflexible work arrangements.

Trigger events may not be always be known unless the manager’s actively engaged with their team members. Events can be far-ranging from getting a degree to not getting a promotion, co-workers leaving, negative news or uncertainty about the company’s future, and major life changes.

Add to this, “out-of-the-blue” job offers and employee anniversary dates. As we know, software developers change jobs every 18 months or so, on average. Some of the best software companies, however, have average retention rates of over 5 years just as chaotic tech startups might have averages of lower than a year.

Retention Risk Metrics:

DriverBehavioral Metrics
  • Sudden decrease of active days
  • Sudden decrease of coding hours
  • Lower push count and frequency
  • Spikes in defect rates
  • A sudden decrease in test coverage
  • Reduced utilization rates
  • A sudden decrease in code review commenting and responsiveness

In a remote work environment, you may have to rely upon metrics instead of observable indicators that you can actually see first-hand at the office. If you still have developers working in the office, or you’re calling them back to the office, you can see these metrics at play with things like:

  • Reduced participation in meetings
  • Long breaks and spikes in personal phone calls
  • Taking more unplanned days off or leaving early often
  • Frequently checking into LinkedIn or Glassdoor
  • Lower interest in pleasing their manager

Replacement costs per venue (job fair vs recruiters) can vary quite widely, but we’re talking interviews, background checks, onboarding time, time for new hires to get up to speed (reduced efficiency), high turnover rates during probation periods, ghosting, snowball effects. We conservatively estimate the overall cost as at least 30% of annual wages or at least $30k per developer. Impact and disruption on delivery schedules could push that a lot higher though.

Behavioral Analytics with a Digital Assistant

The efforts involved in most behavioral analytics tools and processes are mostly manual and time-consuming. They’re well worth the effort, but can be a challenge on the resources of any company struggling to keep key talent. Fortunately, everything that has fallen under the rather broad category of “analytics” can be automated.

Gitential is presently in the process of transitioning its software development analytics from being a dynamic dashboard to an AI-enabled Digital Assistant. Instead of manually digging through charts, statistics, and comparisons, you’ll be able to ask a question and get an answer. It’ll work almost like how you use Google now, type some keywords or ask a question, and get results. The results Gitential will show include the statistics backed by easy-to-understand charts and actionable insights you can begin putting to use immediately.

An engineering manager invest 1-hour using traditional performance analytics to help one developer reduce their defect rate for a very easy 1000% ROI in a year.

With the help of a digital assistant to find more patterns, make the calculations and comparisons, their “Insight Cycle Time” could be 3-5 minutes each. This applies, to help managers 10x their ability to help their developers 10x.

Behavioral analytics benefits most when your entire organization is aligned to helping all “hands on” – everyone – improve their performance. For further benefit, check out 9 Tips for Onboarding Developers as retention efforts should actually start before you even hire someone.

Article Updated: July 13, 2022

Did you like our content?

Spread the word

Subscribe to Our Newsletter

Don't miss our latest updates.
All About Software Engineering Best Practices, Productivity Measurement, Performance Analytics, Software Team Management and more.

Did you like our content?

Spread the word

Subscribe to Our Newsletter

Don't miss our latest updates. All About Software Engineering Best Practices, Productivity Measurement, Performance Analytics, Software Team Management and more.