The impact of Covid on my leadership style

screenshot of the surveytool I build
Matthias Kainer

I’m in a leadership role in a organisation without a strong hierarchy. The company I’m working for is organising itself as a network of highly motivated and capable technologists, working together to create solutions for our clients. In this setup, I’m active in multiple roles: I work directly with teams as a tech lead, I help running accounts as part of the client leadership team, I support multiple people inside of ThoughtWorks, and I am active in whatever position the client asked me to help them.

My leadership mission have been stable over the years, and can be summarised as this: I want to help create successful, self organised teams that are creating outstanding results, providing people a safe space they can use to show their strength and grow if they want to. The goals I have for achieving this missions are separated based on my range of influence, and the role I have when I interact with people. Just as for everything I do, I needed a way to quantify if I am successfully working on those goals, and identify the next feasible steps to take. The solution I found is a relatively simple one.

For years now, I send an anonymous survey twice every year. The questions I ask change occasionally and follow the ideas I found on sites like rework with google. Some questions follow the ideas that we have at ThoughtWorks, others are direct responses to comments of previous runs.

Also, I have the net promoter score (NPS) in there. It’s the question you might have seen in a few surveys already, which goes like this: “How likely is it that you would recommend X to a friend or colleague?”. In my case, X is “working in a team with Matthias”. The number of participants is between 20 and 40 and far from having any actual relevance. Nor am I an expert on analysing a dataset, and I probably make more errors and incorrect assumptions than specialists can bear while trying to follow this text.

But the feedback I received has always proved helpful; people that provided me with answers appreciated the things I took out of them, so I continued with it.

Over the years, for the NPS, not a lot changed. The rating followed a zigzag pattern whenever I changed job or my responsibility shifted and nicely chaperoned my growth journey without providing a lot of value. Comments and answers to directed questions were usually far more helpful.

When I started to look at my survey results for the year 2020, however, I noted that something was off.

The NPS can be a number between -100 and +100. Although it went up and down, it never left the area between 75 and 85. For 2020, the rating was 61 - almost 20 points below the earlier data point.

Development of the NPS from 2017 to 2020

For one exception, the year 2020 has been one of relative stability. I was with the same team/group, and nothing changed for me or the work I was doing. The only change was Covid-19, with the second part of 2020 being a working-from-home thing entirely.

I am aware that a pandemic alone would be a good reason for everyone to be less engaged about what happens at work. The circumstancal environment, however, I can’t change, but nevertheless I might be able to improve the situation for those working with me.

So I decided to check if I could figure out if and what impact the pandemic had on the dropped rating and what I can do to improve it again.

First, I checked the cohorts. The first questions in the survey are providing context on the relationship with this person. It starts with the question about which company the person is employed in, as I work with multiple people from different organisations. For this analysis, I will focus on people from the one I’m working in - ThoughtWorks.

The next question is about the role in which this person interacted with me the most. Possible answers are “Tech Lead”, helping me understand my team impact. “Tech Principal” is used to see how I am doing as part of the client leadership team. And then there are a few that focus around other roles at ThoughtWorks, like onboarding buddy, sponsor (similar to a mentoring role), rubber duck and coffee pal.

For the first two, I managed to improved my rating, while for the latter, the NPS dropped from 100 to 0.

Change of the NPS from a survey conducted in the middle of 2020 compared to the end of 2020

I wondered where this was coming from, so I dug deeper. There is another question I ask: How would you rate my improvement over the last 12 months. If I only look at people that selected “Not working together that long”, the NPS goes down for two of the three cohorts.

Change of the NPS from a survey conducted in the middle of 2020 compared to the end of 2020 filtered by the group of people I worked with less than 12 months

I began to think that this might be related to people I work with that I never met in real life and wondered if I can figure out why they would be less cheerful about my performance. But drilling down also comes at costs - the number of samples got small, and it wasn’t possible to separate a single person from a trend. I changed my approach.

The tool I build for my surveys comes with a few more features that I thought would be cool to code, and now I could check if they would also be helpful.

The first one analyses the directed questions I ask, and for which I provide checkboxes with different positivity. Those questions emphasise essential activities, and I use the answers to see if I live up to my expectations. Examples are “Matthias showed consideration for me as a person” and “Matthias creates an environment that encourages collective learning”, with possible answers ranging from “Fully Agree” to “Disagree strongly”.

The application gives me the numbers of this survey and the ones from the previous years.

Percentage of answers for specific selection in the set of directed questions; comparing all results from the survey from the end of 2020 to the aggregated results of the three surveys

In this graph, not a lot changed. Even though the NPS has gone down, I still seem to have a very positive impression.

Do I ask the wrong questions then?

There are free text comment fields as well, and the application provides sentiment analysis based on AFINN word lists.

The number of comments is small, and the data even less representative than the data before. But still, something interesting popped up. The percentage of very positive words (AFINN score 4) dropped by 5%, AFINN 3 dropped by 10%, whereas the rate of slightly positive remarks (score 1 and 2) increased by roughly 10% each. Changes of up to 5% were common before; 10% was a first.

Percentage of terms with a specific AFINN score in the comments; comparing all results from the survey from the end of 2020 to the aggregated results of the three previous surveys

While the sentiment was very positive overall still, based on the numbers it seemed less enthusiastic. This result ties back to the NPS, which measures a similar thing.

I decided to leave the quantitative data behind and look into more qualitative approaches - I took the question into my one-on-ones and coffees.

In those sessions, I shared the results and asked for feedback and ideas. We can cluster the items that popped out into three topics:

The thing with video calls

I like to think I’m one of the people that stay silent in a meeting. My best meetings are the ones where I didn’t have to say a word because the teams come up with a good approach without me. Independent of the fact that if I might not have taken the same direction, I’m okay with any outcome as long as I don’t have to worry that it will lead to the kind of failure that will negatively impact people.

If I speak up, I try to raise open questions highlighting the issues I see with the current solutions rather than bringing in another. Or open another view on a problem not raised in the discussion but which I feel is essential.

For the rest of the meeting, I stay in the back.

From the feedback I had received, this was one of the things that people highly regarded. People highlighted they feel I give a lot of freedom and trust while at the same time creating the safety that someone with experience will help the team circumnavigate all the potential errors that can have a fatal impact on delivery.

You give us (the team) a lot of trust and just guide us in a direction and you show that basically in every situation across a day, even in meetings with clients.

feedback from the q2-2020 survey

A few people highlighted that with video calls, that strategy doesn’t work anymore. Because no matter how much I’d like to stay in the back, my face is always visible right in front of them.

And those people have highlighted that they look at my face. And try to read it.

I’m not great at keeping a poker face. When I worry, my mouth starts doing things; when I appreciate an idea, I smile. I know that now because I began looking at myself during meetings. That wasn’t an issue in a conference room because staring at somebody who doesn’t say anything trying to read their face is usually considered creepy.

When others were observing me on video conferences, they noticed slight changes in my expressions. Based on what they recognised, they acted. I was pressing them into directions without saying a word.

No longer was I guiding; I was driving.

Perceived as strict and scary

On top of that arose another theme that didn’t help the situation.

With almost 20 years of experience involving multiple companies and two startups as a founder, I’m at a point where I’m very confident. Being no longer afraid of failures but aware I can turn them into opportunities whenever they happen, has become one of my trademarks. Which, I assume, is not always easy for others, especially if you hadn’t had the chance to develop the same self-confidence yet.

On top of that, I have been in contact with most of the technologies that are around. No matter the requirement or problem, I have an anecdote on how I resolved this the last time.

Being perceived as somebody that doesn’t seem to fail and knows everything can make you a person others might see as scary. They wonder if this person will expect the same from them. And they will stop trusting you out of fear of what you might think about them.

I am afraid that you will judge me if I do not do something yet or if I do not know something.

feedback from the eo-2020 survey

In the past, I benefited from the fact that I have another character trait - I am what some call an easygoing person. Rather than being regarded as an impeccable know-it-all, people who worked with me in person attribute me to being calm, relaxed, casual, helpful, and informal.

Continue being funny and light-hearted with mindfulness!

feedback from q2-2020 survey

Looking back at the last year, I can see how a massive face on your screen - that you fear judges every one of your sentences - cannot create a lot of positivity. And there weren’t many opportunities to counter that perception.

By all the measures we have collected, velocity in the teams increased. But at a cost. Back to back meetings on the one end, endless pairing sessions on the other left little time for social life.

Whenever there were spots I had free in my calendar, I helped my two kids with their homeschooling; did laundry; prepared dinner.

But I just didn’t had the time (nor the mental capacity) to participate in any after-work-video-call-get-togethers.

People who haven’t seen me before this episode never really had the chance to get to know me as a person.

Which leads us to the last item on our list.

The difficulty of trusting someone you never saw entirely

In the end, what I feel is behind the identified topics is the simple truth that it is challenging to trust a person you have never seen from shoes to the nose. You lack many things entirely in remote meetings, like body language; specifics in your tone might get swallowed by audio problems. Those are traits you use to create a picture of a person, develop bonds, and connect with them.

And missing all of those things is making it harder to reduce any biases you might have had before or after your first meeting.

In my case, this leads to predicaments with my behaviours - like being overly confident, opionated, scary - that I will have to manage.

At the moment, I’m working on multiple ideas on how to alleviate this.

First, I started looking at myself during calls and avoiding showing more than interest if I don’t want to influence. But this mitigates merely a symptom, not the cause, so another thing I want to focus on is to increase trust. For this, foremost, I’m planning to try out a few things that hopefully help people to get rid of the “scary face on my video call” perception and closer to somebody that you love to work with.

First, I plan to set up a different form of 1o1s. I already have one-on-one meetings, but I’m thinking about introducing quick ones - no more than 7min - with the topic “How’s your day?”. I started the habit of scheduling my meetings as such that there will be a break between them, which I use to get a coffee and relax. I plan to reuse those slots similar to the quick coffee talks we had at work in the kitchen; for short smalltalk.

Another great idea I heard from a colleague is taking a walk together. Even in Germany, a stable network connection is a reality now; taking a one-on-one for a walk in the woods seems like a welcome change.

I also plan to focus more on connecting people that know me with those that don’t. ThoughtWorks is a network-based organisation, and as such, this is beneficial to them by getting to know more people. I hope it will also help me as people that know me can tell others: “he’s not that scary once you get to know him better”.

This, I assume, will scale better than the 7min coffee breaks; although short, there are too many people in my range of influence to have regular catchups with all of them.

Besides, I will have to change something about my perception overall. As long as people will consider me mostly scary or strict, this will become an obstacle to my desire to create self-organised teams. Don’t get me wrong - I don’t plan to become over-ready to be generous or lenient with someone. Having an opinion and standing by it is very important. Software excellence and craftmanship, for example, are topics dear to me and which I will defend with virtue.

I am happy when people are aware of that. Yet, I prefer a state where it is not the first and only thing people link to me.

For this, I’m thinking about changing my meeting style, from silent in the back to more engaged. I assume that a silent person quietly radiating authority is even scarier than someone openly participating. I hope that this will also counter the fear that I would judge by openly expressing that I support their ideas.

And lastly, I wrote this article. To share it with those who gave me feedback, express that I appreciate the time spent answering my questions; that I take your feedback seriously and want to incorporate it to better myself.

And that I will need help in doing so. Let me know if the ideas I have will improve the situation. And when you have other ideas that can help me in supporting you.

Thanks for the time you used to answer my surveys; I appreciate it and continue to use it for my growth journey!