Facilitating the Spread of Knowledge and Innovation in Professional Software Development

Write for InfoQ


Choose your language

InfoQ Homepage Presentations Self-selection for Resilience and Better Culture

Self-selection for Resilience and Better Culture



Dana Pylayeva talks about “self selection” (empowering people to choose their own teams) as a great vehicle to build in happiness, resilience and a better culture. She shares stories from running successful self-selection events, and talks about the “Spooky Questions” game that enables open conversations, helps identify unspoken worries and brings in fun and empathy into a self-selection process.


Dana Pylayeva is an independent Agile Coach and trainer with over 20 years of diverse experience in IT. She’s been on both sides of the “wall of confusion”: first as a software engineer and then as a DBA Manager in low-latency, high-throughput environments. She is the founder of Big Apple Scrum Day conference and a co-organizer of NYC Scrum User Group and NYC Liberating Structures meetups.

About the conference

Software is changing the world. QCon empowers software development by facilitating the spread of knowledge and innovation in the developer community. A practitioner-driven conference, QCon is designed for technical team leads, architects, engineering directors, and project managers who influence innovation in their teams.


Pylayeva: Did you know that nearly 60% of employers have the jobs that stay vacant for 12 weeks or more causing them $800,000 annually? Hiring the right people is hard, but keeping the people that you already have in your organization from leaving sometimes might be even harder. Laszlo Bock, Google's head of people operations, says it best, "The most talented people on the planet are increasingly physically mobile and increasingly discoverable by employers. This global talent wants to be in the high-freedom companies."

What about the companies that you work with? Do you have a choice of selecting the teams? Do you have a choice of picking the project that you work on? Do you have a choice of which technology that you use? This talk is going to be about self-selection, a fundamentally different way of building teams. In the next 40 minutes, we're going to cover all five steps that it takes to prepare the group for running a self-selection for the first time. We're going to get a glimpse of the four steps of the self-selection event, and we're going to explore the impact that self-selection has on your ability to retain the best people, and to improve organizational culture.

My name is Dana Pylayeva. I am an Agile coach and trainer. I live here in New York. I also design simulation and games that help me work better with my teams. I came to know about self-selection from a book written by Sandy Mamoli and David Mole. The book had very detailed steps about how to run it. It also had very interesting and inspiring stories. I loved what I read, I wanted to run it, but the organization I was with at that time was not ready. A few years later, I heard about self-selection again, now at a business agility conference. Amber King was sharing her story about running self-selection with an organization that just runs through the rounds of layoffs.

That is exactly the situation I was in at that time. A large retail company just went through all these massive rounds of layoffs, and you can imagine the teams were just left with holes in them. Management was looking for ways to rebuild the culture, restructure the teams, and they were ready for something drastic. They were ready for experiments, because it was just so bad. Anything you do would not make it worse, it would just improve it. That was a golden opportunity for me to introduce that self-selection. That's how the journey started for me.

Since then, I have worked with these two types of organizations doing self-selection. One of the types of organizations are fast-growing companies that are expanding, they're looking into reorganizing the team so they're able to support more product lines, they're able to retain people who are very talented and looking for new opportunities within the organization. They're able to become a learning organization. The second type are those organizations that I mentioned at the beginning, those that just run through rounds of layoffs. They're trying to rebuild the culture and restructure the teams in a more natural way.

Simple Rules That Guide a Self-Selection

It helps that the roles of self-selection are very simple, because my role in both cases was to be a facilitator to prepare everybody for taking this journey and also hold the space during the self-selection event, to let a group of about 50 to 60 people self-organize into five to seven different teams. The rules are very simple. Teams have to be capable of delivering end-to-end, seek new teams to learn and grow, and do what's best for the company. As you can imagine, there is a bit of a tension between these rules. On one hand we're asking people to think about their future opportunities, their growth, their learning opportunities. On the other hand, we're asking them to think about the company needs.

Let's look at the four steps of the self-selection event and see how those steps help us to leverage that creative tension. These are the four steps, which is finding the card, making informed choices, selecting your team, and checking the team blueprint. We iterate through the third and four steps a few times until we get to 80% of the teams being fully formed. I'm going to walk you through all these steps so don't feel like this is too much or too high level. We're going to go deep in that.

Step one, people are coming into the room. The first thing they do is pick up the card from this table of participant cards. Those cards were prepared ahead of time and now they serve almost as the ticket into the self-selection event. What's on the card? The top portion is very simple. It's just a picture of the person, name, role. Where it gets really interesting, is the middle. In the middle, you have this grid that has team ingredients, and those ingredients are very contextual. They are specific skills that the organization considers as the skills that are needed here to be successful. The green bar represents either the person is interested in learning the skill, so this is "learn," or a person knows how to do it, or he knows how to do it so well that he can actually teach someone how to do it.

The red bar represents the strong aversion to specific ingredients. This is something that that person would not want to do ever, and that's fine. It's good information to have so when we are structuring those teams, we have all these insights into how to best create a team that's able to deliver end-to-end. The part in the bottom is for the teams to enter their first, second, and third choice of the projects as they listen through product owner pitches, which comes in the second part.

The second part is where product owners, sometimes joined by the tech leads or the scrum masters, share their vision about the products that they're building these groups for. They're talking about things like mission, vision, technology that they imagine that will be in this team, what would be the team's culture that they hope to create, and things like that. What's interesting is that those are very short pitches, they're 5 to 10 minutes long. During that time people are trying to almost sell those initiatives to the teams, because everyone in the room - just like you, sitting here today - is sitting and listening to product owners presenting what is that you're going to build the teams for. Through those pictures you get to see the personalities of people, because some people come up with very funny pictures and there’s a lot of humor. Some people are very serious. You get a feel for all the different personalities that you may want to be working with together on the teams. Once they present those pitches, team members have an opportunity to ask questions, and then they will enter those choices on the bottom of the card.

Next step, they're actually going to get up from their seats, and they're going to walk around in the room. They're going to put those cards on the specific posters of the future teams. This is how a team poster may look like. As you can see, there is some information about the different skills that are needed for this specific team, the number of people that we're looking to get into the team. You also see some cards that are there which are cards of those people who just placed their cards into this poster.

The idea is that in the first round, we will not get to the perfect state because in the first round, we're just asking people to put their cards into their first choice. Then we go through steps three and four in the multiple iterations, because this is where we're asking if the team is fully formed. What happens usually in the first round is that we get some over-populated teams and we get some under-populated teams. That's the goal of the next round; to make sure that we let people figure that out. The way how they're figuring this out is, first of all sometimes we have people who don't really care which team they will be part of. There is a special poster for them. It's called, "I can join any team." That's where in the first round, they just add their poster. Then in the second round, when we ask people to rebalance, we first invite people from that poster to join the teams that have available spots. Then we also ask people from those over-populated teams to consider their second choice.

What Could Possibly Go Wrong?

Sounds simple and straightforward, right? What could possibly go wrong? I'm glad you asked because the very first time that I ran the self-selection with the companies, what I didn't expect was that something that was so exciting for me turned out to be very scary for others. That amount of fear that I observed in team members and the management alike, was something that took me by surprise. It reminded me about this quote: "People often get the most afraid just before they're about to step out into creative unknown, into a new possibility."

Running a self-selection pushes the organization into that scary creative unknown even though it might be great for them. But it's scary, it's new, it can trigger many fears. The first one goes back to our school years and triggers gym memories. Standing in line, waiting to be picked, and not being selected. What people don't realize is that self-selection is not like that. You're actually not standing in line, no one is selecting you. It's the other way around. You're selecting the team that you want to go to. Again, this is the first reaction. "Oh my God, that's a gym memory. I can't deal with this stuff."

The second one is fear of missing out, because, "What if I select this team, but this other one is just as appealing? What do I do? Which team will I go to?" Just a bit of that. Another one is fear of the unknown, for management and participants alike. This last one has to do with the middle management: fear of losing control. Imagine for all these years, part of their role was making decisions for others; who was going to work on which project. Now, they're giving that away. They're giving power to people to self-select and pick for themselves. That sounds scary because you're giving away that control, what's going to happen? All those fears that are coming up, and as you're preparing for the self-selection event, you need to think through how to make sure that they have the experience because it is a great thing for people to be able to self-select.

Preparing for a First Self-Selection

It can take anywhere from one to three months to prepare, depending on the organization. There are five areas that as a facilitator you need to be thinking about. The first one is preparing the facilitator because it could be you, it could be someone external, but it's helpful even if you're deciding to run it yourself to connect with someone who's done it before. Even if they can just answer your questions, that's still super-helpful. Preparing the management, because if they don't buy into this exercise, then you can’t do it, so you have to have them on your side. How do you do that? Try to help them solve the real issue. What are they struggling with, that self-selection can help them with this?

The example I was giving from my past was that in that company, there was a new director who came in, and he was given these broken pieces of the teams to reassemble, and he didn't know anybody there. For him, it would be a random decision pulling these people together. He had the problem. Self-selection was a perfect solution for him.

Position this as an experiment. That's another thing that I love doing. Just say, "It's an experiment. If it doesn't work, we'll go back to how we did it before," and so that opens up more doors. Preparing the process - there is a process that's described and it's very simple, but every organization is different. You need to think about what would make sense in our organization? Even putting together those blueprints, you need to define what works for you. Also, if your organization has full-time employees and contractors, you need to make sure that you know who is participating and who is not participating. What are going to be the ground rules for that? Setting up the space - you need lots of space for people to move around. If you can bring food, it always helps. If you can finish with a celebration, you get extra brownie points for that, people love that.

The most challenging part would be preparing participants, because as I mentioned, people have all these concerns. They're also not sure of what's going to happen. What I found out through this experience is that a running traditional Q&A just doesn't work. It does not help anybody, because I get people in the room, I have all the pre-prepared Q&A for them, and I ask them questions. Crickets. No one wants to ask questions when they're in the Q&A. Don't get confused; they have a ton of them. They just don't feel comfortable asking. After the Q&A is over, then you I get asked questions that haven’t been answered; they follow me in the elevator, in the corridors, ladies' room. Anywhere they see me, they would have tons of questions for me, but in a one-on-one setting.

Gamified Simulation: Getting Uncomfortable with the Unknown

Clearly, I had to change the approach. I had to do something different. That's when I came up with the gamified simulation. In this simulation, people turned from being passive listeners to active participants. As they were going through simulation, they were practicing running self-selection, they were doing debriefing, they were doing some gamified Q&A. That way, it was something that first of all took the edge off this experience, because we're just playing a game within the simulation. It was helpful for people to feel how it would be in the simulation, and then ask more questions in the gamified format.

This is how this simulation is set up. I ask people to imagine that they work in this young, first growing company. Our new product requires us to have five different teams that know everything about specific locations in the world, because we're going to build these mobile apps to support specific locations. For this purpose, we're going to assemble five teams, and we're going to let them travel to that location, all expenses paid. All you're selecting is just a specific city that you want to go to. In the simulation, people create their own cards, thinking about their learning preferences, their choices for what they want to learn and what not. Also, what is the city that they want to be part of?

This is the picture of one of the simulations that we have done. People go through product owner pitches, they go through the selection, they go through all the negotiation, just like they would do in real life, except there is no limit to imaginary perks that you can offer people to consider their second choice. Fun fact: every time during the simulation with the teams or in conferences, there is one city that always wins. Hands down. That city is Kyoto, everyone wants to go to Kyoto for some reason, I don't know, but it is beautiful.

I'd like to experiment here as well and take a turn with you. Imagine now that you work in this amazing company and that company is forming five new teams. You have a choice to go to Kyoto, Chicago, Zurich, Cape Town, or Dublin. Again, all expenses paid, you're going to be there for one month, and your goal is to learn everything about the local culture, or try all the different local restaurants, try the different entertainment, whatever is there because we're building that beautiful app. Think about for a moment, which travel team would you join, and what would be your second choice? In the first round, you also have an option to differ the decision and select, "I can join any team."

Here it is, this is our first selection. As you can see, we have some cities that are more popular than others. Chicago didn't get any votes so far, we have 10% of people who can join any team which is good because this is our first round. What we're looking to get is we're looking to get equal distribution of people in every city or close to equal distribution. Good to know that in the first round, it didn't get there because it would be just magic if it did.

Now, what we're going to do for the next round, we're going to vote again. The way we're going to vote, is if you voted for Cape Town or Kyoto, you can just vote the same way. If you voted for, "I can join any team," in the second round, that's your time to make a decision. I encourage you to think about Chicago because you're going to get some awesome benefits to those people who go there. People who go to Zurich, imagine any perks that you can get. You have it in this simulation. I encourage you to consider the second choice because our company does need to have people in Chicago. Let's vote again.

Let's look on this second one. Again, we're still not there yet but you got the idea that we got some people agreeing to go to Chicago. Then we've got some people moving their team. You get the point. When we do it with the teams, we actually walk around the room and we have all those posters so it's a little bit closer to the actual self-selection event. But because I was told this was going to be rows of chairs and I can't ask people to move around, I had to come up one step. Hopefully, it gives you a bit of a feel for the choices that people are making, and then some people are open to changing those choices.

Simulation Debrief with Spooky Questions Game

When I run this with the actual teams, what we do after the simulation is, we go through a debriefing. In this case, Zurich was the most popular city. What is Zurich in our portfolio of projects? What is the project that's going to be so popular? How would we handle the over-population in that project? What incentives can we offer people for them to consider a second choice? Because in a simulation everything is possible, but in the real life, you may want to know what's available. That's a conversation that we're having, and it's helping people to build that first transformation memory that they just went through, the simulated event. Now, when they go in the real one, they have something to think back on and relate to.

Another thing that we do is, we run this Spooky questions game. The spooky questions game is a game that I created using a traditional Q&A, but turning it into a game, to creating a set of cards - one of them with questions, one of them with answers, and then asking the teams to play the game to help them answer their own questions. We're going to try a set of Spooky questions right now and that's why you have those printouts on your chairs. The idea is that there are not enough printouts per person and that's intentional. You would have to come into groups of two or three because that's when you're going to be working with those cards together.

What you have in your hands, those are answers. I'm going to give you three Spooky questions, and your goal is, using those pre-prepared answers, try to find how to answer those questions. Talk to your partner in this game and think about what is the right answer for these questions?

Did you find the answers? The hint is the numbers on the Q and numbers on the A, they match. If Q13 has the answer and it's A13, that's a perfect match. That's just the hint. This is theater sitting so I didn't want to cut out the cards and actually play them as the card game; when I do it with a team and they're on roundtables, that's what we do. We actually play the game. You can shuffle those cards; you can pass them around, it's definitely a little bit more engaging. But still, the point is the same. You're going through the answers together. You have your questions that are pre-prepared, but the trick is that I also have another card and that's a joker card. That's the card where you write your own questions.

Before you start the game, whatever questions you may have, you write them down. Then, as the team is playing, you don't know whose question is which because you're shuffling them around, and as a team you're answering those questions. That is a much safer way for people to raise those concerns and ask questions that they were afraid to ask in the Q&A, the traditional setting. It's also great feedback loop for me as a facilitator, because I collect those questions afterwards. I go through all of them and analyze what is this team or this group of people struggling with? What are they unsure about in terms of self-selection? This is one of the ways to gather information as you're getting ready.

I just wanted to share some examples of Spooky questions that I collected. People have questions "What if I'm selected?" "What if I'm not selected?" and "If I'm not selected, am I going to lose my job?" That's a big concern that people have, even though to me it's a positive thing, "Yay, let's do it," but people have those concerns. "What if my new team or I don't have enough skills, what do we do?”, "What if I join the team with team members I don't like?" And many other questions.

What I'd like you to do now is to imagine yourself in a self-selection in the organization. On the armchairs you have those index cards. I'll ask you to pass index cards to your neighbors, people who are around you so you should have at least one index card per person. What you're going to do next is, you’re going to think about if you were in the self-selection, if your organization was going through self-selection, what kind of questions would you have? We're going to take that minute to write those questions down.

Now, I'm going to ask you to stand up. You're going to find a person in this room that you haven't had an opportunity to speak with before. You're going to share with that person those questions that you wrote down. For the next two minutes, you're going to discuss those questions. Do you have the same questions? Do you have different questions? Can you help each other answer those questions based on what you've heard so far? Maybe it's in those sample answer sheets. Let's experiment, let's try that.

Now, you're going to take your pair or trio - if you have a trio, that's fine too. You're going to join another pair or trio. For the next five minutes, what we're going to do is we're going to now discuss the questions that you discussed with your first group together. We're going to bubble-up the most difficult questions that you can't answer yourself. I want to take those questions and answer them for everybody. Get together with a bigger group, four or five people, or six.

Let's hear those difficult questions. Remember, it's one question per group.

Participant 1: Our question is what is the optimal size group to begin doing self-selection? Can there be too few people, is a thousand people too many?

Pylayeva: My experience has been 40 to 50 people, and I've heard people running much larger self-selection events. To me, when you're just starting out, it should be a group that can self-select into four teams roughly. Then, if you're thinking about scrum size teams, that's around nine people max. So that's 36 people.

Participant 1: How would you manage a team of 100 people who need to do self-selection? Do you break them up and do two separate ones, or how do you approach it?

Pylayeva: I would still do it as one, but I would not start with 100. If this is your first time doing self-selection, I would start with a sub-set, or do a simulation, maybe just a trial-run, maybe do a hackathon where it's just one day, you're self-selecting for that one day. Get experience running for the first time a small group, and then do it with 100.

Participant 2: One of our questions was how do you balance out the skillset for your team, so you don't get someone or a group of people with one skillset?

Pylayeva: You have the team posters, that big sheet that describes how many of what kind of people you need. You need specific people with these kinds of skills to be a cross-functional team that's able to deliver end-to-end. That's how you create a blueprint ahead of time, and during self-selection you're letting people to select into it based on their skillsets.

Participant 3: The question we came to was, what can I do if I don't get the team I really want to go to?

Pylayeva: The idea is that you typically don't do it once and for all. This is something that a lot of organizations do on a continuous basis. The frequency is something that people decide what works for them, but the idea is that if you know that you're with this team for maybe three months only, the next part of negotiation could be that next time when there is a self-selection, I want to be on the team for sure. That could be part of your negotiation when you're selecting your second choice.

Participant 4: How long is too long to stay on a team? If you do this, then the person is on the same team over and over again, and I have experienced this.

Pylayeva: Well, I'll give you a typical answer, it depends. Whatever works for you. If the person is reconfirming, recommitting to the same team over and over again, is there a way to encourage that person to consider other opportunities in terms of learning? If that person is only focusing on this part, on this technology, then effectively that person's not growing. For that person and for organization it’s not good. Is there a way to find how to encourage that person?

Participant 4: We probably needed to incentivize it more and more up until a point where making it requires becoming the only... our number right now is two years on a team, you're forced to move.

Pylayeva: That's good information because that's something you could build in into your roles of self-selection, to say that if you have been with a team for two years, then you have to move.

Participant 5: What if no one picks a certain project or team, say maintenance or production support, versus the cool new upgrade?

Pylayeva: Remember, do what's best for the company. As a company, we have to keep the lights on, we have to do that. That's where the negotiation comes in, and how can we make a deal that this person will be on this maintenance project for this period of time? The next time, someone else gets to be in that position for some specific incentive. That way, you can negotiate.

Participant 6: Does this mean you end up being volatile?

Pylayeva: Is it better than having 100% of people being volatile?

Participant 6: Sometimes that happens to me since I have the key skill for the production support. I always end up being volatile, and keep signing this card.

Pylayeva: If you have the key skill for production support and you're the only person who has that skill, how good is that for the company? That's actually not good, because if you win the lottery, then the company is in trouble. It's in the interest of the company to have more people picking up that crucial skill so we are able to be more resilient.

Participant 7: What's the cadence at which you reselect? If I, as an individual, feel like I would like to join a different team, then that just happens, or are there fixed cadences based on projects or sprint schedules?

Pylayeva: That's a great question because there are so many things that go into play. You want people to commit to a specific team for a period of time, because as you know, there is that way in which teams are forming, storming, norming, performing. You need to give your teammates a chance to become a high-performing team. If you're switching every week, probably you're not going to get that chance. There are definitely bounds of how long a person can stay on the team and how frequently you can do it. One way to sync up the selection is with the release cadence, because sprint cadence might be too frequent.

In the last minute, I'm going to tell you to give it a try. Remember, freedom to self-select into your own teams can be exciting and scary, but breezing into your fears will bring you a step closer to the culture of engagement, experimentation, and the learning. Give it a try. Take some time to prepare. I know some people had questions about why there are four steps, also five steps. That's just the numbers that I came up with, this is what I've seen work. When you run the self-selection event, start with a small group. Don't start with a huge number of people because that's the easy way to fail. You'll learn a lot from it, of course, but try with a small group. You can even start with the minimum viable experiment. Try the gamified simulation, try maybe the hackathon. Give it a try. If you want to learn more about it, there are good places to start.


See more presentations with transcripts

Recorded at:

Aug 26, 2019