On the bare necessity of psychological safety
medium.com
medium.com
Previously, this would just have been part of team alignment. In a business culture that eshews black-and- white thinking, the zero-sum state of "safety," should be replaced with something more dynamic, like risk, because then it is something manageable you can reasonably engage with. Accepting the principle of psychological safety implies that it's not just a (unfalsifiable) fire alarm lever to pull. If anyone told me they felt "unsafe," it would be difficult not to make every effort to avoid them, and I think any lawyer would say the same. US law has this concept of "fighting words," which this concept is not directly, but appealing to safety is absolutely a legal and political threat, and it is not normal professional discourse. Since I am not a psychologist, I am not equipped to deal with psychological safety, and this is an HR issue.
Imo, we should really look at reframing "psychological safety," as something more manageable like, "burnout risk," "initiative risk," "working trust," or something else less formal.
To be effective, a team needs to be able to communicate freely, and each individual needs to be able to speak freely. But there’s a lot of power dynamics at play, not just in the team, but external to the team.
This is also why minorities are especially cursed in work environments: it’s not just that some folks might actually have prejudices against you based on your identity, it’s also that some folks might have prejudices against you based on your identity, and so you watch what you say and/or try not to be too assertive or contentious and/or wonder if some poorly worded comment was intended as a slur or not...and the smart manager is also cursed, because they may feel that they too need to be careful not to be too assertive or contentious, lest they draw attention from HR or a lawsuit.
Successful teams require mutual respect and brave communication. “Psychological safety” is a term that feels correct to me in the abstract, but very wrong as an applied concept for the workplace.
I think “cultivating mutual respect and open communication” is probably a better framing.
> Since I am not a psychologist, I am not equipped to deal with psychological safety, and this is an HR issue.
And yet, the way you choose to communicate and interact with others contributes to how safe they feel. You aren't formally educated in psychology, but it's still your informal responsibility as a member of society, to generally not cause psychological harm to those around you. Including when at work. If you don't take that responsibility, then you become the HR issue.
> zero-sum state of "safety," should be replaced with something more dynamic, like risk
How is risk any less zero-sum than safety? If you're going to claim the only states are safe/not-safe, why doesn't risky/not-risky work the same way?
> we should really look at reframing "psychological safety," as something more manageable
I think you're putting the phrase on too high a pedestal. To paraphrase the article, psychological safety is feeling that asking for help or admitting failure will not be punished. That's it.
> burnout risk," "initiative risk," "working trust," something else less formal.
Those all seem more formal to me, though I think "working trust" is at least in the same ballpark.
If you are avoiding someone, you feel unsafe. That is what it is, no reason to avoid that word.
The only way I can see psychological safety effecting the eventual outcome of Google+ is someone who, being psychologically safe, would have said right at the beginning, "this is stupid, don't even try it".
Nobody's suggesting this is a panacea that'll make your unviable idea suddenly viable, the article is about how options for working that could lead to success get cut off when PsySafety isn't present on the team.
How many line engineers at Google let alone a random tech company can question a VP/director's pet project? Heck, how many middle managers can? Psychological safety should mean you should be able to call BS and be okay with others calling your BS (in a non-inflammatory way), but in reality it means stroking others' egos, especially of those in leadership.
>It is easy to see the faults of others, but difficult to see one’s own. — Buddha, Dhammapada
Indeed, like all of Buddha's teachings, it also applies to the author of course.
And this is engineers at Google we're talking about: usually people with advanced degrees from hard-to-get-into colleges and years of experience along with very specialized knowledge. In other words, very hard to replace people. How must more difficult must it be for people who could be replaced tomorrow?
Maybe 2005 Google, but not anymore. A good bachelors degree or enough YoE at most companies gets you the interview, and then the interview result usually does the rest. Those are the "line engineers" that move on to other faangs or unicorns when psychological safety becomes an issue.
Typically it's the people from the hard-to-get-into colleges, or with years of company tenure, or with specialized knowledge causing the psychological safety issues.
This is an extraordinary claim that requires extraordinary evidence.
I thought Google just hired recent grads that play the leetcode game.
You work on technical tasks for other people, who make sure your work is ok, and your supervisor makes sure you are OK.
That person is your advocate in all things that fall in the tense employee-employer tug-of-war. Inside the "group", there is psychological safety. On "tasks" there is often less of it, but you always have your group.
Or at least it should be that way.
The elephant in the room is that there has to be a level of tension, of slight mutual discomfort, of a bilateral transaction present in employer-employee relationship. It needs to be very dull and implied but it has to be there to make sure incentives are aligned and objectives are consistently met.
Tenured public service and unions show what might happen when this inconvenient element is missing.
If you won’t disagree and commit that’s a different story.
I have had some co-workers or higher ups that probably found it inconvenient and didn't like me (in fact strike the probably - I know for a fact). Yet it has so far not resulted in me being fired, demoted etc. and in fact the opposite has happened. I've been told to never shut up and that they keep me around precisely because I don't.
My luck might obviously run out at some point but so far, from my first internship where I just didn't do the stupid task they gave me and instead replaced their awful sendmail+pop3d setup with a full featured postfix and imap/pop and spam filter combo (well I showed the boss on my last day that I had built it on my computer and what it could do. He offered me a part time job as their network/sysadmin to do it for real on the server and to migrate all their users over) it has worked out. I'm not on 'brilliant jerk level'. Way not brilliant enough and just below jerk ;)
He almost got fired recently. He didn’t, in part because a bunch of us stood up for him, but it was dicey.
The reason he almost got fired is because he’s very expensive to manage.
That dogged unwillingness to never shut up, along with more than his fair share of brilliance, is great when he’s digging into a thorny, ugly problem that most would prefer to pretend doesn’t exist.
But when there’s a meeting with $3k/hour in fully loaded cost worth of people in it and he’s going on and on about his pet issue that’s a problem. When every decision you make has to first be delivered to the team and argued in detail to this guy, that’s exhausting. To his credit he’s convincible, but still. Since projects have to be carefully evaluated for fit before being assigned to him that makes him harder to use well. Having to unruffle other employees’ feathers all the time certainly isn’t fun. And so on and so forth.
All this means he needs a very strong manager, both in a technical sense (because without that credibility no one will be able to manage him) and in a people skills sense (both for him and for the people he comes into contact with). Such managers are themselves are difficult to find and expensive to employ. To keep him around and aimed in the right direction means a disproportionate amount of one of those people’s time is going to a single employee.
I’m not trying to convince you of anything. It sounds like whatever you are doing is working for you. But maybe this provides some food for thought.
To make another example of where I personally can't shut up: if I see someone propose a stupid rollout plan that's got a real good chance to eff things up royally, then I will say that it stinks and why. That doesn't always mean doing it in the big 3k/hour meeting like in your example. And I'm not talking about a 'pet issue'. I can obviously be overruled and that's fine. I've brought it up and have it black on white that they were warned and they made the decision anyway. Fine by me. Don't blame me afterwards. I won't do unpaid overtime later to fix your shit.
But I sure have seen way more than 3k/hour spent in meeting upon meeting after the warnings weren't heeded and it did explode in their faces.
Actually being unable to focus oneself reasonably well without the help of a manager sounds almost a bit dumb to me
I'm surprised you're writing "really bright"
> if I see someone propose a stupid rollout plan that's got a real good chance to eff things up royally,
To me, you seem to be the opposite to those "really bright" people, in that you're helping others, sometimes the whole company?, focus on the right things, and avoid doing dangerous things.
Those others, need help with focusing. Whilst you help people focus correctly?
Most people are not equally bright in all departments :)
Yes, they're bright guys, I stand by that and they do awesome work of what they do. When left unchecked, they just focus exclusively on what they love and think makes the most sense. This may not always fully align with the overall company goals customer wise. And obviously you're getting only _my_ viewpoint of this, not theirs.
They do remind me of myself more than 15 years ago, telling my dad about Linux From Scratch and Gentoo and how that's the only way to be. He was obviously right when he said that no, they don't have the time for that at their place of work and I declared him insane (in my head) because obviously, that was _the only_ way :) It definitely made me learn a lot of stuff, so for myself at the time it was the right thing.
> in that you're helping others, sometimes the whole company?
I try I guess. Sometimes it's really hard, because they don't want to be helped and/or actively seek to introduce policies that are detrimental to the company overall (not talking about the "bright guys" above in this case, e.g. the rollout plan I mentioned was the business people, not us tech guys). And again, my viewpoint only that you are getting here.
That's the kind of psychological safety I have.
Curious to learn others' experiences, but I've never seen this in my career, ever. I've seen _peers_ yell at each other, but never seen an employee slip "out of character" with their superiors, at least publicly. And if it isn't happening publicly, there's no psychological safety, is there?
Most people (including myself) have been conditioned to bring their "work self" to work, not their "true self". As such, there's very little chance of "forgetting the nature of the relationship" even in heated moments.
I am not really one to challenge all that directly normally. But sometimes that slips.
So when I was reestablishing my personal identity in my late 20s, I just never made a separate “work self”, and it’s been a deliberate decision. I regularly talk openly with my superiors and make it known that I’m very opinionated but I’m also fallible and open to learning from my mistakes. This kind of openness and trust is absolutely essential at the executive level; the real secret is that none of us really knows what we’re doing and we’re all making it up as we go too.
Doing this a few times with the right people will help you learn when asking questions is helpful and when you should get out of the way and follow orders. A good leader has the level of humility to know they’re not always right.
In fact, this ability to “fight back” effectively is what gets you promoted above a line manager level. And in my opinion is why white men are so overrepresented in senior levels of companies: the psychological safety is there for them in the rest of their lives so they can have it at work, and it’s not there for others so it’s a lot harder to build the muscle on the job if you’re always on guard at home.
This is amazing self-awareness. I wish I had it earlier in my career.
>And in my opinion is why white men are so overrepresented in senior levels of companies: the psychological safety is there for them in the rest of their lives
This is interesting, but I don't entirely agree. I have seen both sides as an Indian male who has worked both in India (privilege) and SV (no privilege). It's not a matter of "fighting back" or "being open", it's just in-group mechanics at work. There is a tacit understanding among people in positions of power about who is the "in-group" that doesn't need to be expressed or acknowledged even. For example, in India, even as a junior engineer I would be invited to important meetings whereas more experienced female engineers wouldn't. If I'd simply shown up and done the bare minimum, I would have quickly risen up the ranks, IMO. I had the exact opposite experience in SV, being seen as the "out-group" of the "worker bee" class by default.
I'm disputing that thesis based on my experience. You don't need to be vulnerable or "be open" at all - you just need to be part of the in-group. I know "Senior Director of Engineering" who had no clue what the hell was going on. They never spoke up in meetings, or did anything useful at all - it was clear to everyone they added no value whatsoever. Turns out they were just floating from startup to startup moving up in seniority in the process. Presumably, nobody bothered to question them during the interview process as it was assumed they were "competent" based on their background and past positions.
It was actually kind of interesting, since I thought as you did before that.
There are always consequences for over sharing, the obvious answer is don't overshare.
That talk is not actually meant to matter, it's just cant...
How can any human feel totally psychologically safe in a world with climate change, cancer, random attacks of violence, earthquakes, etc?
The answer is that you're thinking in black and white terms about a concept that is a continuum. Even in employee-employer relationships, with a good manager it is entirely possible to have a reasonable level of psychological safety, just like it's possible to have a reasonable level of physical safety while driving a car.
Then again, it's reasonably straightforward to measure physical injury and come to some agreement about acceptable risk. Psychological injury is far more subjective.
And if the group is small, even if you try to make the answers anonymous, maybe you still would be able to figure out who wrote what, and they'll realize this, and write mostly nice sounding things
Anyway, asking and talking about it seems like a good idea. Maybe also asking them to read the article being discussed here :-)
Speaking my mind does not increase the speed of climate change, the chance that I might get cancer, or the probability of earthquake. Therefore, it is easy not to think about those things while discussing work.
Obviously, "total" in this context means in the workplace. Also, psychological safety essentially means "ability to express yourself and make mistakes without the fear of negative consequences" - it has nothing to do with shielding yourself from cancer, climate change etc. Ironically, the Buddha, who was quoted in the OP grew up with "total psychological safety" as you defined it, and you know where that led him.
> just like it's possible to have a reasonable level of physical safety while driving a car
Again, this is not a comparable analogy at all. If you're unable to in no uncertain terms say "this project is going to hurt users" out of fear of getting fired/looking like not a team player, that is not even a "reasonable level" of psychological safety.
We're talking about software engineers in big tech companies, right? They're extremely well paid and would have no problem finding a new job.
All of them, as long as it's done internally and not jerk-like. Doesn't sound like you know much about Google culture.
Really? Why are y'all working on FLoC then? :-)
It's not an average SWE's job to make product decisions and to steer strategic initiatives. There are people whose job it is to do specifically that.
Think just for a second what it would be like if every one of be more than 100,000 employees could bring a product plan to a halt because they had reservations.
What's in your power is to ask tough questions (people do).. To make your opinion heard (people do). To change projects if you wish (something Google makes easier than any other company). And to take risks and make choices in the execution of your own job duties. That's psychological safety.
You may find this hard to believe, but the people working on FLoC have chosen to. They could easily find a different project if they wanted to. It may be disappointing you you that the average SWE isn't nearly as ideological as the average HN commenter, but it is what it is.
Employees should have psychological safety to voice concerns within their domain, but that's not a license to start backseat driving upper management. It's one thing for a line engineer to raise concerns about their team's choice of frameworks for a project or the chosen architecture of the system. It's a different issue entirely if the employee is simply "calling BS" on company directives or cynically referring to initiatives as "pet projects". From a managing perspective, there is a stark difference between people who voice concerns reasonably and those who simply like to complain.
The good team members raise their concerns with an open mind, seek more information to understand why certain decisions were made, come prepared with alternative suggestions, and most importantly are willing to disagree and commit if they don't get their way.
The bad apples do things like "call BS" on initiatives without understanding the whole story, aren't interested in learning more, don't have any constructive suggestions, and tend to drag their feet or spread dissent through the ranks when they don't get their way.
Psychological safety doesn't mean employees get a free pass to be disgruntled or push back against what they're being paid to do. Questions, concerns, and alternate suggestions are welcome and should not be punished, but at the end of the day employees must be willing to commit to the chosen direction and get on board with company initiatives.
I'm still trying to learn to keep my mouth shut. Speaking up is detrimental unless it conforms to the existing group (really leader) sentiment.
Actually, I tend to be the one advocating for the grey area solutions. I think in the West we have too much emphasis on one of two opposing solutions being the complete and correct answer, when the best option is probably some amalgamation of the best parts of both ideas.
The other issue is how do I tell if it aligns with what they want? There's so much salesmanship in the leadership talks at large companies. I've brought in suggestions that clearly aligned with stuff stated in those sorts of meetings, only to have it basically ignored. With some of it I was able to find out from friends that the real goal behind the scenes was completely different.
I generally only bring up problems if there is a solution for it. Otherwise it's just complaining, which I still do but not to my boss.
- your boss can fire you at any time without cause or notice
- your project can be canceled on a whim
Since these properties are generally foundational and immutable (and usually codified in corporate governance as well as employment law, at least in the US) "psychological safety" cannot exist in any such organization for any rank and file employee (executives may be protected somewhat by contracts and golden parachutes). The only thing that can exist is a hierarchy of fear and power, and that is exactly what we see.
Would that not be more "safe" for everyone involved? That's but one tangible, measurable way we can increase psychological safety in organizations. Your claim of immutability seems to lack basis.
This is rediscovering what has been going on (or at least held as the ideal) in the military for quite some time.
It's a two way street of accepting a contrary opinion on one side, and continuing to work as a team when that contrary opinion is rejected. I know it's popular among some to view the military as brainwashed morons, but it is a conscious decision to execute orders that one disagrees with. Otto von Bismarck Quote: “A bad plan that is well executed will yield much better results than a good plan that is poorly executed.”
Management is foremost about solving the problem you are accountable for. In development, that will likely mean that success is when the team ships something useful to the business.
In an unsafe environment, members of the team are afraid to act and speak. The working relationship devolves to "tell me exactly what you want done". Teams can ship like this but managers need to overperform and it sucks for everyone.
In a safe environment, manager and team have a shared understanding of what the goals are, and we work flexibly to acheive them. if something doesn't work, learn from it and try something else. Collectively, we still have to deliver. For sure, we can all be safe and still fail. Safety needs to improve the odds.
The manager question is "how do we best ship?" What dynamic best predicts team success for a given goal? The bet is that a safe team will perform better at solving the typically squishy problems that stand between you and shipping well.
My belief is that safe teams perform better by allowing more people to engage constructively and critically on any one problem (more traction), by learning efficiently (low latency), and by demonstrating work incrementally and frequently (more cycles). This is failing fast, failing better, and converging well.
One additional point of "safety" is the question of "how do you want live?" As manager or team, what do you want work to be like? People flee harsh teams. How do you succeed then?
I worked at a company where vulnerability and mindfulness were championed above all else.
It went exactly the same way as Mao's [Hundred Flowers Campaign][1].
So much so that I am convinced it's something deeply ingrained into human nature.
Feedback was welcomed, and psychological safety was guaranteed.
Initially people were rewarded for being vulnerable and opening up with feedback in public and anonymously. It ended up as a witch hunt with the CEO trying to guess whose anonymous feedback was whose and admonishing people's anonymous feedback in public meetings, and then several employees being fired after the CEO went around spreading rumors about the employees and building a consensus that things weren't going well because of these employees - things that everyone else could see were clearly false.
It was incredible to watch, and when it happens it creates the ultimate chilling effect, and rapid loyalty to the leader on the witch hunt, but closes everyone off, and makes everyone parse their words and basically never say what they feel needs to be said. You don't push back or tell them they are wrong or what they are doing is inappropriate, because you see how vicious they are to dissenters. It was an amazing psychological journey to go on though, and I learned a lot about human nature, and a first-hand understanding of how some events of history probably played out psychologically amongst a population.
People who obsess over, and push mindfulness and vulnerability and such are usually those that are over-thinkers and with psychological problems that led them there in the first place.
The problem is that the people in power don't realize that they actually can't handle to hear a lot of the honesty they are asking of people.
So I guess my point would be that, psychological safety is important for eliciting true feedback, but even if you are the one in power and offering it, you cannot trust yourself that you will be able to handle it.
Everyone should just be more chill and easy-going instead of trying to drive formulaic new-age initiatives.
> Awesome! Not awesome that we have this problem, but awesome that you brought it up. That‘s how we build psychological safety and that’s how we build a successful company!
This gives me chills haha. So robotic and corporate. I just wish people would chill!
The way psychological safety works is that sometimes one new person is all that it takes to destroy the safety. Every time you add a team member, you roll a die. The chances of continuous good outcome decrease exponentially with the team size.
Even this is not completely unavoidable. Sometimes the existing team is strong enough to resist the new person, if the new person turns out to be incompatible with their previous relationship. And sometimes the new person leaves, and the peace is restored. But this is rare.
Specifically, I have seen a situation where a new junior developer with somewhat toxic personality was treating every existing team member like an idiot. This did not ruin the psychological safety of the remaining members, because being attacked all at the same time made it obvious they were all on the same side and no one took the attacks seriously. (It would have been worse if the new member instead focused his attacks only on one or two of them.) And because it was a junior developer, his opinions had no real impact, so people around him felt uncomfortable, but not actually threatened. Finally, it was the new guy who decided he had better options than to work with idiots, and in a month or two he moved to another project. The atmosphere within the team quickly returned to the original state.
A much worse situation was in a different company, where the new member was maybe only 1/5 as toxic, but he was a senior developer recommended by the management as an expert, and stayed there for a year before moving as an expert to another company. The original peace was never restored, and at the next problem, people started quitting.
> The way psychological safety works is that sometimes one new person is all that it takes to destroy the safety
I suppose that's a reason CEO handover is considered risky -- rolling a die with the whole company?
> different company, where the new member was maybe only 1/5 as toxic, but he was a senior developer recommended by ...
I wonder if he got feedback about his behavior? I guess that maybe it wouldn't have had much effect but maybe
People who consider themselves smarter than everyone else usually don't pay much attention to feedback from those supposedly less smart. We tried to explain where he was wrong, but he thought we were just making excuses.
Can total transparency in a large organization, damage psychological safety
I wonder about the Linux mailing list, the whole world can read. I wonder if some good ideas never get posted, because there're too many people reading.
Hmm, if there was a way to post ideas semi anonymously, and take credit for them, only if they turned out to be well accepted
We (some of us) use shorthand (most of the time) because if we didn't, communication would be more difficult (in most cases).
I agree with most of the rest of your comment.
Whose idea was that? It's odd if the initiative was from the crazy CEO
> Everyone should just be more chill and easy-going instead of
Imagine the managers said "be more chill and easy-going". I wouldn't know what they had in mind.
The things in the article, though, included some precise and actionable steps. Which I'll make use of, where I work.
But with such a CEO, maybe the company was a bit doomed in any case
A maxim worthy of framing on wall.
> The problem is that the people in power don't realize that they actually can't handle to hear a lot of the honesty they are asking of people.
This too!
People feel "psychologically safe" because they have confidence. They have confidence because they have a track record of getting things right. They might be wrong now but they have enough credit to risk being wrong. Successful teams have people who speak up because those people are confident (likely due to a history of success).
In any case, in the event that you speak up and are wrong, there's no way you can be insulated from "formal nor informal consequences of interpersonal risks". If someone says something that I know to be wrong, it makes me think less of them. Will I be rude to them? No, but if it keeps happening I'll pay less attention. Kind people will give you plenty of chances but only stupid people will give you infinite chances.
The perspective here, where we should aim for "psychological safety" is minimizing one kind of fail case (people not speaking up because they're afraid) and maximizing another (people speaking up and being wrong). It isn't obvious that you should make that trade-off.
> That's stupid, it won't work due to X, we're going with Y
or a
> Thanks for the feedback, but this time we're choosing Y due to X.
One is going to leave the worker happier than the other and feeling safer despite the boss thinking the worker is wrong. Unless you can measure the counter-factual (often not) you will only guess as to which was better.
What are your concerns with solution A such that you think solution Y is better? Help me understand so I can explain what you're missing. Or, so I can see what I'm missing.
Often the biggest challenge is team dynamics. An introvert on a team might feel his psychological safety reduced by what he perceives as relentless criticism my an extraverted teammate. However, said extrovert might view psychological safety as being free to express his opinions as he sees them.
I've had to manage for PS at several organizations. It doesn't mean that everyone is free to say/do whatever they want whenever they want. It's team-oriented.
It takes a lot of work to establish if you don't have it. It's incredibly easy to lose if you abuse it in any way. But it's an awesome way to work.
It's weird, there are so many of these pieces and they all sound the same. Is there a template that everyone uses who wants to get into personal brand building? "Start with a direct quote. Then set the context with some sentences. Now throw in a snippet from some study you've read. And use funny images!"
1. Hire someone.
2. Pay them at top of scale
3. Give them 1 years salary upfront, cannot be claimed back. (Effectively one year rolling tenure)
4. Repeat for all employees
5. Loudly and clearly state the problems facing the business.
6. Stand back
I think the idea is that financially secure people are more willing to speak their mind. Getting salary upfront means everyone has small scale “fuck you” money. The problem here is that most people don’t actually have a reason to care about their employer’s problems, so not speaking up is often just a strategy of least resistance.
I was a part of that once. It was a complete trainwreck. Never again.
Being humble is often about taking the initiative to express mistakes, errors or things you did not know or thought you knew.
In an optimistic mindset, “psychological safety” boils down to feeling secure in one’s job and in control of one’s project. When people feel secure, they free up cycles to creative, sleep better, enjoy each other’s company more, etc. when they feel in control they have the energy to constructively argue.
But when rubber hits the road most engineers do not have one of either security or control. At a startup they may have some limited product control but there is no safety. In a large company they may have more security but no actual control of the product. Either way, to promote PS and get a more engaged team, a manager has to downplay the reality of the situation. This leads to management allowing complicated pet projects, rewrites, and feedback meetings that go nowhere all for the sake of appeasing egos. Engineers tend not to be overly people-savvy but we catch on eventually.
I think it’s no better than the “we’re all a family” rhetoric. Just another way of trying to eke out more work for nothing in return. My last employer signed all employees up for an app called “happify” where we would get daily challenges to pop bubbles with stressful words on them or write praises to our coworkers. That strategy was so bald-faced that I had to kind of admire it.
So why are checked-out engineers bad? They are worse than bad, they are counter-productive. They push shit code because who cares? They don’t question obviously bad designs because they know there’s no point. They quit or move on leaving the team in churn-hell. But these are symptoms not of psychology but of reality. Spinozas Ethics is a lovely look into how psychology and reality are just two sides of the same coin, and you can’t fix one without fixing the other. As stated above, the solution is engaging work, but most work is not engaging. So let’s just be more realistic, ok?
Every manager (as well as a few colleagues) has tried to get me to come back so far. A couple have recommended me to lead the tech side of their friend's (well-funded) startups long after I left.
I wouldn't ever do it again though unless it was explicitly asked for. Stressing about getting fired in the initial weeks after getting hired (or during a boss change) just isn't worth it.
Being able to be your "true" self is the observation, not "be happy". If your true self is a neurotic mess, "psychological safety" on a team would give you the space to do that without feeling judged or that negative consequences would arise (informal or formal) by acting that way.
You can improve "checked-out" engineers feelings by creating a way for them to express their concerns without worry that they'll be punished for it.
Those toxic teams I'm sure you've been part of, where everyone felt shitty and nobody felt connected to the work? They didn't have to be that way by leveraging "psychological safety", even if "reality" was immutable (also, it's not immutable but we can get to that).
My main point in writing the diatribe above was to deter the type of pleasant but dull middle manager who thinks that all they have to do to solve their teams’ problems is psychologize them. I’ve actually noticed this as a larger cultural trend: people blaming their own minds or the minds of others for what is an actual real problem in the world. If I can help clear the wool from a few low-level programmers eyes to help them see that they aren’t so different from an accountant or the fry cook at their favorite bar (and that this is not a bad thing!) I feel like my job is done.
The trick is to find people like you, get you engaged and feeling like you can actually fix the shit that's stopping you from being productive, while also getting you to give space for people to fuck up now and then, because people fucking up and feeling okay with that sometimes leads to really great insight when instead of fucking up they do something great.
Basically, "safety" is letting people feel not-terrible for taking a risk and having it blow up in their face. That should be allowed on a team, and Google's research seems to show that teams who allow people to fuck up without making them feel bad about it tend to make better shit than teams who punish people for fucking up.
...to a point. If all you do is fuck up, that also violates the safety, on the manager's side because then they're not feeling safe letting people take risks.
What it comes down to is a culture where decision making is transparent, questioning the status quo is encouraged and no one person holds the power to make / break a product. At the team level you’re always going to have good managers and bad managers, but with the right culture the bad managers blast radius will be diminished. A good manager can’t do a whole lot to overcome a shitty culture.
This is a utopia though. Every manager has their own manager right up to the company board. If you have to let people go, then how can you say you are not going to fire the guy who is the most checked out? If the person cannot deliver you cannot guarantee they won't lose their job.
It boils down to you really can't trust yourself to not be biased to those who are more open but more negative.
Simple, you cannot guarantee "they won't be punished".
Most checked out is not the same thing as can't deliver. In a well-functioning organisation one would hope that managers would assess people on how much they were actually delivering rather than how involved they sounded, and would have support from the structure above them in that.
I am in a situation where I actually have near absolute safety and nearly no control over the projects. It was even worse in 2020 regarding control after being thrown into a toxic project by a CEO who always touts the motto "no a**holes".
Well when pitching for the biggest budget in the history of the org reality is what counts. Not nice mottos.
So I agree. Everytime I hear this bs about being a team/family/safe place I just think: STFU and pay me. I do my work. Nothing more, nothing less.
The observation is if everyone feels comfortable disagreeing with one another, the project is more likely to succeed. That's it.
This is a productivity tool to make marginal projects more likely to tip favorably, and to keep good engineers around even when they're on bad projects, and hopefully make some bad projects less bad by giving everyone room to suggest improvements.
I would hazard a guess that most of the misunderstanding around the term is due to its ambiguity. I think it's so bad it’s counterproductive actually but that's subjective. It really feels like what they are trying to say is that everyone feels personal agency in their decisions, actions and communication, but the term evokes a concept of being free from threats or risk.
I would suggest that neither of these create high performing teams, they just tend to reduce the pathological behaviors that impede them. They still need to get past the observations of Andy Grove 'When a person is not doing [their] job, there can only be two reasons for it. The person either can’t do it or won’t do it; [they are] either not capable or not motivated.' After all, someone sleeping at their desk day after day could actually be experiencing maximum psychological safety.
It is both.
My decision to stay or quit is a cost-benefit analysis, and feeling happy or unhappy is a part of the equation. High employee turnover usually means lower productivity, because knowledge is lost and new team members need to learn about the project from scratch.
Maybe it would be helpful to use the term other than “safety”? Is there a synonym for a psychological safety that would get cross what you’re thinking about better?
Trust. Psychological safety is (real) trust between members of the team and the people who impact the team (e.g. managers or executives or whoever). It's trusting the people around you so you feel free to be yourself, to propose off-the-wall but potentially terrible ideas without being thought less of, to be open about flaws and failures, to try to stretch yourself beyond what you think you are capable of because you trust that the people around you will support you.
It isn't having a secure job or even a happy one. I was once on a team with a shitty director and severely understaffed in a very unsexy, low-paying industry. I did not enjoy my place in the organization. But we had psychological safety on our team. In our scrum team the tech lead/manager was amazing. He protected us from above, he trusted us, he was honest about the situation, and he had a passion for building a great team and growing his people. When someone made a big mistake in prod, he taught that person how to correct the mistake as an opportunity for growth while the entire team supported them (but didn't do it for them, because we trusted that person). The job sucked and within 2 years the entire team had quit. But not a single person even thought about leaving until that team lead left and the culture of trust left with him. Psychological safety is related to happiness in that teams with it tend to also have happy employees - but it creates happiness, not the other way around.
Psychological safety is often hard to have when the job is unstable (it probably feels unstable because of a lack of trust somewhere), but a startup is a great example of where you can have both trust and an unstable job.
"Happy" is a hard term to define, and I don't think you need to be happy to feel "safe". Another way of describing "safety" is "the ability to make mistakes/ask for help without negative consequences". You may feel "unhappy" that you made a mistake, but you won't feel judged for it if the people around you are comfortable with you taking risks.
You can be deeply unhappy, personally and professionally, and still experience that shared comfort with risk taking on your team, and benefit from what Google's research shows, I think. Perhaps you're all comfortable with the risks because the project is going so poorly and you need to throw some hail marys out there to try and get out of the bad situation. The focus is on everyone agreeing that making mistakes is expected and acceptable, because risks also come with reward.
I actually really like that example, where you're on a doomed project, and the natural "safety" that forms with the team who all knows it's a doomed project anyway, so why not take a few shots at something wild? The space you create in that environment where, "we were going to fail anyway might as well throw some shit at the wall to see what sticks" is exactly what I mean when I say "safety". Projects that have that attitude apparently tend to be more successful. Imagine having that level of comfort with your team and also working on a successful project!
This is not true. I don't need to be happy to want constructive arguments. In fact, the more unhappy I am, the more I need that constructive argument about what is bothering me - to deal with situation that makes me unhappy.
Whether I go into it depends a lot on whether I think/feel it is safe for me to open argument or join it. If I think I will be punished for saying stuff, I don't say that stuff, because I am not dumb.
And I can also be unsafe psychologically and subjectively happy. I can have interesting tasks, fun, I won't be expressing my opinions and there will be risk unsolvable situation will arise. But until then, I will be happy.
You know what? Most of the time they are right. But it's still important to have a culture where everybody feels comfortable presenting their own ideas and critiquing others - not because we're going to discard the sr. staff persons architecture in favor of the interns idea very often, but because the process is good for the ideas, good for the product, and good for the team members (whatever level).
"But when rubber hits the road most engineers do not have one of either security or control."
Well said
Does it always have to be dishonest? "We need to implement a bog-standard CRUD app for XYZ, so it's a good opportunity to try some new technology stack / make a really polished codebase / ..." ought to be the kind of conversation that you can have.
The point is that all members on a team may believe the work is meaningful and interesting, but still be unproductive due to group dynamics that penalize certain types of behavior that would constructive if they were allowed. It doesn't matter how meaningful or interesting things are if a co-worker or manager shuts you down every time you present a differing opinion on something.
The process is exactly the same as when on tech suddenly everyone talks about fluent api one year, then about other buzzword, just for those then slowly dying.
Unfortunately the adaptive thing to do here is go along with it. If a friend tells me they are being gaslit, I’m certainly not going to side with the gaslighters.
My favorite example of "marxist" style insults is claiming that someone lives "rent free" in ones mind. Admittedly, I find this uniquely insulting and if anything I've got to give credit to zoomers and the other forerunners of todays "wokism" for adding more interesting insults and accusations to our linguistic zeitgeist.
Lawful Neutral (Apple)
Lawful Evil (Google)
Chaotic Evil (Facebook, Twitter)
I attribute a lot of this to weak social safety nets and at-will employment in the US. I constantly feared that I would be fired for some statement, personality clash, or mistake (or for no reason at all, which is entirely legal)... so I kept my head down and was lightly abused (frequent overtime without pay, tolerating verbal abuse, etc).
I was living paycheck to paycheck and without a job I'd rapidly lose housing and healthcare. It took me nearly a decade to attain financial security while racking up psychological damage.
Between the lines, I'm reading that a good social welfare system, eg health care and unemployment money, would be good for the companies -- easier to reach higher psychological safety, then, more good ideas and healthy critique, better decisions