Safety Check

by Steven M. Smith · 40 comments

You have heard repeatedly that an agenda is a vital ingredient to a successful meeting. But little is ever heard about safety in meetings—the environmental variable that determines whether people participate or merely observe. How do you measure safety? What actions are available to leaders for creating a safe meeting environment?

First published in the Jan 2006 Issue of CrossTalk: The Journal of Defense Software Engineering.

He is wearing his traditional garb: dark suit, white button-down shirt, red tie, and black tasseled shoes. The glare off his wire-rimmed glasses makes it difficult to see those steely blue eyes. Harry Fox has all the right moves, and his quick climb up the management ladder proves it. He is arrogant and ruthless. People who oppose his ideas pay a price. And the payment is extracted when they can least afford it.

We are both participating in a problem-solving meeting. Well, that is not quite true: I am observing and Harry is talking. He just stole the floor from Jim King a few minutes ago by talking louder than Jim. I hate that behavior. Jim looks dejected. Harry continues to dictate his ideas about how the team should solve the problem. I realize that Harry missed three crucial facts, which will cause his solution to fail.

Should I share the facts? Wait a minute. Harry does not like to be corrected. He wants to hear only the facts that support his position. Harry is connected all the way to the top of the company. I am connected to the people on my team. I hesitate. Wow, that is totally uncharacteristic of me: I am known as someone who speaks his mind. I look over at Harry. He has taken his glasses off and is moving them rhythmically up and down as he talks. Although what he is saying does not make sense, it sounds authoritative. I feel my gut twisting. Is it anger? No. It’s fear.

Harry concludes his speech. There is a pause. If I want to speak, it’s time… I say nothing.


The omission of crucial facts and opinions happens in thousands of business meetings every day. If people do not feel safe, they are not going to say anything. And you will have no idea about what you missed.

Too often the participants who are the most vocal assume that everyone feels as safe as they do. This assumption is wrong more often than not. But it is rarely ever tested.

You can help increase the safety of your meetings. Collect data about conversational safety. Share it. Interpret it. And decide how to respond to it. These actions will open the opportunity to transform your meetings. For instance, you will create the opportunity to discuss and take action on items previously not discussable such as who was or was not invited; what is and is not on the agenda; and how the discussions will or will not be processed. I have experienced the power of this transformation many times. You can too.

Collect the Data

Inform everyone that you will use a secret ballot to poll the participants about their safety to speak freely. Poll people with the following question: “How safe is it for you to fully share your ideas during this meeting?”

Write this question on the board or a flip chart. Clarify that the ballots are not identified, just a number on a slip of paper. Expand on what fully share means by listing some controversial ideas that were not shared at other meetings that would have made a difference.

An unsafe environment causes participants to share fewer ideas and to carefully filter the ideas they do share to be sure they are safe. Poll people for the information in Table 1.

Level Description Comment
4 Secure Everything is discussable without filtering
3 Safe Almost everything is discussable without filtering
2 Neutral Most things are discussable without filtering
1 Dangerous Many of my best ideas are not discussable
0 Treacherous Most of my best ideas are not discussable
Table 1. A Safety Gradient

Pass out a ballot—a small piece of paper, Post-it Note, or note card—to each participant. Ask everyone to write the number corresponding to their level of safety on the ballot using the numbers zero through four as defined in Table 1. My experience is that some people will, regardless of the instructions, write a decimal number. Simplify things for yourself by informing everyone that all the ballots will be rounded so that the results fit the range of the gradient.

Ask them to cup the ballot in their hand when writing the number so that no other participant can see their rating. Stress to everyone that you do not want anyone to share their rating with anyone else, regardless of how safe they personally feel. Again, emphasize that only you will see their ratings. Have the participants fold the ballot in half and place it in a container, such as a hat.

Share the Data

Ask a participant to help you build a histogram of the poll. I suggest that you use a flipchart so there is a hard copy of the histogram to use when you write up the minutes of the meeting. Pull each ballot out of the container one-by-one and read the score to the person building the histogram. Stuff the recorded ballot into one of your pockets or put them in your briefcase so no one else can or will ever see them. Note that you are not only revealing how safe people feel — you are also building safety by checking numbers in a way that reinforces safety.

Table 2 shows an actual histogram built during a requirements-gathering meeting that I facilitated at a large manufacturing company.

Level Description Number of People
4 Secure XXXXX XXXXX (10)
3 Safe X (1)
2 Neutral XXXX (4)
1 Dangerous XXXX (4)
0 Treacherous (0)
Table 2.: The Histogram From an Actual Safety Check

Interpret and Respond

Ask the participants, “What is your interpretation of the histogram?” A manager in a requirements-gathering meeting said they needed to start trusting each other. His management colleagues vigorously echoed his belief. And his colleagues had a lot more to say about the importance of trusting each other. I let this discussion continue for 10 minutes and asked, “What cluster of people on the histogram do you think is offering the most advice?” The room fell silent. The people who felt the safest realized that they were doing the most talking. They realized that the people who felt threatened were not talking.

Telling people how they should feel does not work. And, in my experience, people know that as a fact, but forget to put that knowledge to work. It helps to give them a gentle reminder. Ask everyone, “How do the participants who feel completely safe help the participants who feel threatened?” The answers I have heard in meeting after meeting can be summarized in two words: care and listen.

During a manufacturing meeting, people did start to care and listen. The participants slowed down and asked each other questions. Most importantly, they were okay with moments when no one spoke. I believe that silence is a gift. It shows people that you are ready and want to listen. And, in the case of a meeting, silence demonstrates that the group is ready and wants to listen.

These changes made a big difference in the requirements meeting. The discussions were deeper. The enriched conversation enabled the discovery of requirements that would have been invisible to them. They were more effective together than they had ever been.

Other Methods

Another method that can help create safety, especially in large groups, is to let the participants build the safety guidelines for their meeting.

Split the participants into small groups. The ideal size is a triad—three participants. Ask the groups to 1) introduce themselves to each other, and 2) create a set of guidelines for conducting a safe meeting. Give them a few test cases to ponder. For instance, someone starts blaming someone else, someone tells an inappropriate joke, or someone dominates the meeting, and so on. Let everyone know that they should not limit themselves to the test cases. You want them to share any guideline that will make the meeting safer.

The hope is that the discussion will help remind people of what they already know about safety, and remind them to practice what they know. Just as importantly, the hope is that a connection with a small, manageable number of people will increase safety.

Have each small group introduce their members and share the safety guidelines they created with everyone. You will be amazed at the wisdom that people have about safety. Gain agreement from everyone on which guidelines to accept. Remind them that the guidelines are theirs rather than yours. If someone violates a guideline, you will call them on it.

Ask the group to monitor your facilitation and to inform you if you allow any deviation from the agreed-upon guidelines. When someone mentions a deviation, treat it with the utmost care and respect. It is the ultimate demonstration of the value you put on safety.

Final Thoughts

Although the methods I discuss are especially valuable for setting the right tone for organizational improvement efforts or multi-day meetings such as a project retrospective, they are also valuable for reoccurring meetings. The key is to expose, explore, and respond to feedback about safety. If followed, the feedback will take the group in the appropriate direction. Feelings about safety will change so it is a wise investment to have a process for periodically exposing and responding to issues about safety.

Regardless of the method used, you can never be absolutely certain that all the participants feel safe. If someone would have asked me how safe I felt during the meeting with Harry Fox, I would have voted neutral or safe so that Harry would not find out.

The best that you can do is to solicit and respect everyone’s ideas. The leader who models appropriate behavior in meeting after meeting is constantly renewing and enriching safety and productivity.

Be a leader. Care. Listen. Model the behavior you want.


A special thank you to Jean McLendon for making me aware of the importance of safety and how to measure it; Jerry Weinberg for suggesting that the number one is not nearly as evocative as zero for connoting the absence of safety; and Esther Derby, Don Gray, and Jerry Weinberg for sharing their feedback about this article.

{ 17 comments… read them below or add one }

Michael James February 23, 2012 at 11:19 pm

The first time I encountered this type of safety check was as a participant a couple years ago when Petri Heiramo ran a general retrospective for our training team, and the last time was yesterday afternoon when I ran a Sprint retrospective for another friend’s Scrum team. I would not want to leave this out of a retrospective, especially with a new or unfamiliar team.

I want to underline Steven’s advice “Stress to everyone that you do not want anyone to share their rating with anyone else, regardless of how safe they personally feel.” The people who already feel safe are protecting the anonymity of those who aren’t by being discrete about their own votes.



Steven M. Smith February 24, 2012 at 4:55 am

Thank you, Michael, for sharing your experience performing a Safety Check. My experiences since writing the article continue to confirm that without safety individuals in group discussions will unconsciously conspire with each other to avoid talking about the most serious problems. It doesn’t have to be that way.


robert murphy October 15, 2012 at 4:14 pm

how can you implement a safety check in a geographically dispersed team where anonimity is void as everybody has logged onto a central location and there is no option for an anonymous vote?


Steven M. Smith October 15, 2012 at 7:05 pm

Hi Robert, I suggest using a service like SurveyMonkey to do the Safety Check. It’s a means for providing people anonymity. Best wishes on a successful safety check.


George Dinwiddie October 24, 2012 at 9:31 am

The January 2006 issue of Crosstalk is now found at

Steve, this is a really good article. The only thing I have to add is that (as I mentioned at the safety exercise only brings out the reported level of safety. If people feel extraordinarily unsafe, they may not be willing to admit that fact, particularly in situations where someone says, “We’re an open team and everyone feels safe.” I love your question, “What cluster of people on the histogram do you think is offering the most advice?”


Steven M. Smith October 24, 2012 at 8:06 pm

Hi George, Thank you for the feedback. It made my day. I tried to say that people may report feeling safe to hide their true state. But I gather it didn’t come across clearly… Thank you for the correct link to the Crosstalk issue.


Venkatesh March 15, 2014 at 12:08 am

I agree with you George. I have come across teams where they are so scared to even admit that they are unsafe.

Also, if it is a small team then it becomes even uglier as the “trouble maker” could be easily identified


Steven M. Smith April 11, 2014 at 7:51 pm

Hi Venkatesh, If the “trouble maker” is easily identifiable, remove them from the team.

Warm regards,

Bill November 3, 2013 at 2:20 am

Hello Steven,

Thanks for writing this article, I learned something.

I do need to ask, wouldn’t it be a violation of the anonymity of the balloting to later ask about ‘who is doing the most talking’? Because this would likely identify who voted “secure” or “safe”, leaving the rest of the group identified as one of the other three options. It could also come across as a betrayal of the safety guidelines one just tried to establish.

Maybe a solution is to ask those who are busiest to answer that for themselves, inwardly. Or just ask to hear from those who haven’t shared yet, or something.

What do you think?


Steven M. Smith November 3, 2013 at 7:49 am

Hi Bill,

Thank you for the feedback.

I gave your question some thought this morning. I don’t believe it’s a safety violation to ask who is doing the most talking. I could only speculate about whether the person answering the question feels safe or unsafe. Feeling safe doesn’t prevent a person from seeing the paradox of their behavior and commenting on it. Feeling unsafe doesn’t prevent a person from saying, “The people who feel the safest.” I maintain my anonymity regardless of whether I choose to respond. I doubt the people who feel unsafe will comment, but again, like everyone else in the room, I can only speculate about how that commenter feels about the safety of the meeting.

Wishing you safe meeting environments to learn and grow,


Nick Zdunic November 26, 2013 at 5:21 pm

Been wondering how to be safe without necessarily asking for safety directly. Thx and I will incorporate this going fwd.


Steven M. Smith November 28, 2013 at 10:45 am

Hi Nick, I don’t know. Wishing you success.


Bryan January 13, 2014 at 12:11 pm

Thanks for this article.

Does it make sense for me to be a little hesitant to implement this check on small teams ( ~ 5 or 6 people ) since the less people there are, the less anonymous their response will be?


Steven M. Smith March 6, 2014 at 9:50 am

Hi Bryan, The method has less value for a small number of participants because of the reason you state. But discussing safety and assumptions about it would benefit the participants.


Dennis Grimes June 30, 2014 at 8:45 am

This was very informative and I will do this at my next team meeting. Being a formal Marine and Drill Instructor, I’ve been told that when you 1st meet me. I come across very aggressively. So, when I come to a new team as a manager, it takes awhile for them to understand my personality. Using this Safety Check will help me and my team members work better together.


Steven M. Smith July 5, 2014 at 7:34 am

Wishing you success.


Asti April 20, 2017 at 3:49 am

Thanks for sharing this article,

Nice article about safety.


Leave a Comment


{ 23 trackbacks }

Previous post:

Next post: