Harry Vos

How I used feedback to analyse my work behaviours

Sticky notes on a screen. Mostly out of focus. One reads, "I'm inclusive in decision-making and onboarding". Another, "Engaging everyone in the team"

In late February, I found out my role was being cut. So I’ve been looking at product manager job descriptions again. I’ve been questioning how rarely I came across desired behaviours in contrast to technical skills. From what I could remember, my past feedback from colleagues was mostly about my behaviours. If I was right, I pondered to what extent recruitment processes allow me to highlight my most widely observed strengths. I’ve been questioning this at each stage; application, interviews, personality tests, tasks etc.

To gain deeper insight, I analysed all the written feedback I had from colleagues and managers. Spoiler alert. Yes, most of my feedback was about my behaviours, not technical skills.

Besides the analysis confirming my hunch, I’m sharing my findings for two reasons. First, I want to hear your thoughts on assessing behaviours when hiring. The other is that I’m curious to see if anyone thinks they need a product manager with these behaviours.

Behaviours observed over seven years

Harry Vos with a colleague. They're sat at a table, both smiling and making eye contact.

Photo by Frederik Højfeldt Nielsen at the Feminist Futures Copenhagen Hackathon.

I collated feedback from 2017 to 2024, so there’s some variation in my behaviours over time. Around 90% of all the feedback is what I’m doing well and 10% is where I could be better.

I’ve attempted to mark what I see as technical skills with an asterisk. However, the delineation between behaviours and technical skills is fuzzy.

Doing well

Around 40 mentions each:

Around 30 mentions each:

Around 20 mentions each:

Could be better

Sticky notes on a screen. Half out of focus. Visible "Learn to be comfortable with not trying to solve every problem". "potentially being able to say no to things - maybe we're doing too much?". "I should publicly share my work more". "Going forward don't be afraid to discuss with leaders how to communicate it more widely (given the internal and external interest)."

Around 10 mentions each:

Around 5 mentions each:

What I learnt from analysing my feedback

I’m using the constructive feedback to improve. Publishing this blog post is a good start for “share my work publicly more often”. Also, I hope sharing this can help hold myself accountable for change.

My colleagues and managers observed my behaviours more often than my technical skills. So I think my behaviours are my strengths at work. As a result, I’ll experiment with my approach during recruitment processes. I'll try to highlight my most widely observed behaviours.

Analysing my feedback also boosted my confidence. This helped after they let me go. I’d recommend this as an activity for anyone who is having a wobble. I’ve shared my method at the end of this post.

What’s next

Thanks for reading. Please comment or reply, as I want to know what you think. How much weight do you think behaviours should have compared to technical skills when hiring? This could be about product managers or similar roles in tech.

Also, feel free to get in touch if you need a product manager with these behaviours. I’d appreciate you sharing this post with people you think might be interested.

Reply on LinkedIn

Reply on Mastodon

Notes on how I analysed the feedback

  1. I collated feedback from colleagues and managers. Most feedback was from forms that I asked people to use. Some was anonymous and some unprompted
  2. I split out sentences that mentioned different behaviours or skills. This meant I could group them separately
  3. I used the affinity mapping technique to synthesise from the “bottom-up”. This led to clusters of feedback
  4. I wrote a provisional sentence to summarise each cluster
  5. I read through everything again to check if each piece of feedback fit better in another cluster
  6. I merged similar clusters, aiming for insights while avoiding losing important nuances
  7. I updated the cluster summaries to reflect the new clusters
  8. I counted how many bits of feedback were in each cluster
  9. I excluded both clusters of two and clusters where feedback was from only one or two people. This was to focus on observations confirmed by multiple sources

Confirmation bias may have affected my analysis. I started with a hunch that most of my feedback was about my behaviours. To limit the bias, I reviewed the feedback again. I looked for possible clusters of technical skills. This resulted in one new cluster, “I give articulate presentations”.


Previous: How to run an employee demographics survey