How to Transition Your Employee Into a Better Role

Kevin Perko

Head of Applied Research / Data Science at Scribd



I recently hired an employee who I felt was promising, but who struggled immensely and had a hard time adjusting for a number of months. I was determined, nevertheless, to get them to become a successful and high-performing employee that I believed they could be. After I changed the type of work they were doing and expanded the challenges, they suddenly started to blossom.

Actions taken

I initially tried to provide additional one-on-one coaching which didn’t bring the desired results in terms of unlocking their potential and improving the performance. Then I decided to put them on different tasks and track their progression.

First I tried to put them on more business-oriented tasks that included intensive communication with business stakeholders. However, they were not making any significant progress and I later realized they lacked the specific business context that would enable them to communicate smoothly with business stakeholders. This particular task required mastering nuances captured in business communication that they weren’t familiar with. My first experiment obviously failed and created frustrations on both sides. At that moment, I thought I was stuck.

Then, I tried having them work on some smaller, well-defined projects and I was glad to notice that they were doing decently. I wanted to step up the difficulty level and introduce some harder projects. Yet, I was cautious because when I tried doing that with business-related tasks that had put us back.

Finally, I decided to assign them a much harder machine learning-focused task. It was a huge leap in complexity and open-ended ambiguity. To my surprise, they were ready to handle it with minimal guidelines. The more arduous challenge helped motivate them and unlock their full potential.

At that time another person was working on a machine learning recommendations model that was taking too long and didn’t bring satisfactory results. I asked my promising employee to go through the code and they managed to break it down and optimized it in a very short time. Nothing made sense to me because it was not a gradual progression of complexity but a huge leap forward, entirely non-linear. This person excelled at ambiguous, ill-defined projects, working on the deeper technical, ML projects that included deeper programming. Once given more onerous challenges they began to shine.

After my experience with this particular person, I had to reevaluate my approach. I would still start with smaller tasks but I wouldn’t hesitate to move them to more complex tasks even if they struggled with smaller tasks. To put it simply, it seems that the two are not correlated.

I would have them work on multiple projects before drawing any consistent signals or conclusions. I would need at least three projects to understand where my employees are in terms of their abilities and performance. One project is almost never enough to make any final assessment because there are too many external variables that could confound the signal. However, the length of the project assigned matters greatly; three days would certainly give you less signal than three months.

If there is something that gives them a hard time, I should uncover that and precisely define it. Is it the technical or communication part or is it about the deliverables or timeliness -- I would try to extract that out and work with them to identify knowledge/skill gaps.

Lessons learned

  • Separate different areas of performance when you are onboarding someone. Establish with the greatest detail how they are performing in all of those different areas -- are they performing well in terms of communicating with stakeholders or doing the work on time.
  • Add gradually a more specific type of work. Learn about their background and what they have done in the past and expose them to different types of tasks and levels of complexities.

Be notified about next articles from Kevin Perko

Kevin Perko

Head of Applied Research / Data Science at Scribd

Leadership DevelopmentCommunicationOrganizational StrategyDecision MakingCulture DevelopmentEngineering ManagementPerformance MetricsLeadership TrainingPerformance ReviewsFeedback Techniques

Connect and Learn with the Best Eng Leaders

We will send you a weekly newsletter with new mentors, circles, peer groups, content, webinars,bounties and free events.


HomeCirclesCircle Ideas1-on-1 MentorshipPeer GroupsBountiesBecome a mentorChangelog

© 2023 Plato. All rights reserved

LoginSign up