What Google Learned from Its Quest to Build the Perfect Team
nytimes.com
nytimes.com
I think how well a team works together is sometimes more important then how much knowledge you may have of some arcane framework.
This is one key advantage of startups, actually. Just take your dream team, and co-found a startup with them. Find the business that your particular group can excel at. The more this becomes a viable alternative to just taking a job, the more we'll be optimizing the matching of great teams to big opportunities.
1 - A team can succeed in one environment but fail in another.
2 - If the team does fail, you have a lot more rebuilding to do.
3 - Teams that are independent of company can have different loyalties. The principal-agent problem [0] gets a lot worse if an entire team is being held hostage.
4 - Many times you're not building the project team from scratch. You start with people who have succeeded on other projects.
[0] https://en.wikipedia.org/wiki/Principal%E2%80%93agent_proble...
I've also watched an entire frontend team quit and go work at a different company because, well - they were a pretty well-knit team, but more loyal to each other than to the company...
That kind of thing happens when a big company acquires a small startup, for instance (and it's the main reason I always see acquisitions/acquihires failing spectacularly) - again, all anedoctal...
Sometimes these worked because the acquiring firm had no expertise in a particular area. The biggest impact moving in teams has is that you are familiar with a lot of the business customs already - you have a support team in place. Your shared network can grow a lot more quickly. However, with a liftout, you have enormous pressure to succeed; there's less 'onboard' time since you should know how to work in your team already.
[1] http://press.princeton.edu/titles/9128.html, Chapter 7
I think team-matching service as you described would be something a lot of people from diverse fields will be interested in?
I'm sure they worked better as a team than each one in a different company now.
"RPS: We’re really interested in your view on the independent games scene. I’ve just spoken to the Portal team and discussed Narbacular Drop and the job offer, and their shock and delight at finding themselves in that position…
Gabe: Something that gamers should probably understand is how important it is that game teams stick together. No matter how good a job a team does the first time they make a game they’re going to do a much better job the second or third time. There’s just so much value in a team having shared experiences to draw on, and my reaction looking at these kids was that they had done this fabulous thing. I go to all these trade shows and see all these tedious, derivative, lifeless games, and these kids had done something that was better than 98% of the gameplay I see. The idea that they wouldn’t work together again was a tragedy. They needed an opportunity to work together and ship a full-on game. If they were able to do that exciting a game the first time, then it’s nothing to what they’ll be able to do in the future. It turned out to be a really good idea."
https://www.rockpapershotgun.com/2007/11/21/rps-exclusive-ga...
Whole interview is great.
The business model for app companies and contract software firms comes to mind. The team is really a small business itself and signs contracts with other companies to build things. Is there something different to these that you'd see? Or is it the idea of a team being hired on directly by company after company vs building to a contract?
The problem of hiring individuals to join the firm still exists though. Although I like the approach you propose more than the "dump a bunch of resumes" approach.
Never the less, one of the terms is from psychology, one of the terms is from pop culture and may not overlap completely
Psychological safety is less normative: It's about being in a space where it's safe to voice an opinion, and feel that you're heard and valued for your contributions. It's a term that can apply in a very culturally diverse team.
Because being open-minded is their enforced conformity.
And if the leader of the team doesn't respect a member, he won't listen to the member, no matter how long the member gets to talk.
It's Feb. :D
> I'd much prefer reading about companies trying to improve their working environments than about whatever incremental new product feature they're launching.
Google has set the right precedent for the tech-industry, and I am grateful for them having done that. 'Cause other big tech-cos like Amazon sure don't give a damn about employee well-being (not just warehouse workers, but engineers too).
The pattern is "If a junior woman can't code, let's promote her as team lead; If a junior man can't code, let's teach him to code" - and this is wrong because both sides have everything to lose with this behavior.
Basically, give people your full attention when talking to them, and avoid primal displays of dominance or subservience. Mostly the attention part.
Could you elaborate on this? How do you resolve significant differences in height?
Another important bit is that if someone is sitting down, you should sit down next to them. If they're at their desk and there's no chair, squat or kneel.
Leave escape room.
Whether it's a bar or a cubicle, a woman will (in general, according to my wife and other female friends) be aware of being pinned in. For those not thinking of it, it's perfectly natural to stand at the end of a bench seat or in a cubicle opening to talk, but this gives them no "escape" from the situation and increases tension. Simply standing to the side of that escape path makes the situation less tense.
On the generally-male side of things, (drawing purely from my own experiences) I definitely notice when tall people are close to me, but if they stand a little further back I don't really notice their height.
E.g. NBA players who huddle around their coach
Also, don't sit straight across from them, police interrogation style. It's intimidating, especially in situation like an interview. Sit at the orthogonal side of the table.
(Disclaimer: This is my opinion, not that of my employer.)
If you have to pick, your team sucks.
I suspect a confusing of "harmony" as a situation in which people don't speak up out of not wanting to hurt someone else's feelings with "harmony" where everyone is comfortable speaking their mind as they know their comments won't be taken as an attack or unnecessary criticism.
The second definition of "harmony" is actual harmony. The first is anything but.
I'll just quote what the Rust team said about their code of conduct, because it's spot-on: "The Rust community doesn't subscribe to the notion that there's a dichotomy between intellectual discourse and kindness."
That's pretty much it. If your team considers those two mutually exclusive, you're headed for trouble. It might not be today or tomorrow, but you will pay for it long-term.
edit: or if it does hurt their feelings, that they'll be able to express that safely, and you'll all be able to resolve the conflict.
A thick skin is necessary, but the point of psychological safety is that it must be normal and expected for someone to make mistakes and to be wrong. It happens inevitably. How teams handle it is the important part. The most successful teams I've experienced are those where someone can say "yeah, I messed up, I'll fix it" and the team sees this as a necessary part of a team's long term function.
That means nobody on the team says "I told you so," and that team members can disagree intensely without making it personal or emotional.
The big problem with "without making it personal or emotional" is that the people most guilty of making disagreements personal or emotionally charged are also least likely (in my experience) to recognize that they're instigating. These people are toxic without realizing it themselves.
I definitely don't mean that. More like, there are always tradeoffs. There is cost/benefit in being truthy just as there is cost/benefit in being harmonious. Each has it's own type and you get to choose which make more sense for the situation, team, etc.
I don't think you can "have it all" so much as you can choose to set the tradeoff point somewhere in the middle rather than at the extremes. For example, in the Rust code of conduct example, it sounds like they have decided they want a nice mix. This will cost them though. Maybe the cost is as small sending $some_multiplier_above_1 the amount of emails between the team because things are worded more indirectly to not offend and therefore more disambiguation has to take place. If their team instead committed to be as direct as possible, the could spend less time on follow up emails but now they'd probably turn of or drive away some good folks from the team who decided they didn't like all the directness and send of urgency.
Personally, I like a nice mix. I think it is just worth being clear with myself it is a mix and not two independent things where you can have an infinite supply of each.
I believe you may be misunderstanding something about the idea of psychological safety. I can't think of any definition of "harmony" that has anything to do with psychological safety.
Makes some sense. Team meetings are not where anything gets done after all. If team members are engaged with their peers, they may invest in team goals and perform better.
Issue: they correlated team meeting metrics (talking time; socialization time vs problem solving time etc) with team success at achieving goals. They could have used behavior outside the meetings (communications per day; social interactions after work) and predicted successful teams that way. It doesn't mean that "Meetings that go like this indicates good teams"? I.e. changing your team meetings to look like what good teams do, may not improve your team at all. That might be put under 'cargo cult management'.
You know, I think it's worth pointing something out about the word "safe" given the sense it has been used in lately. In this case, it means people feel safe to express dissent. Not "safe from dissent", as the term is often used nowadays.
Of course the two usages shouldn't be overlapping much anyhow. It shouldn't surprise anyone that a team where people actually attack each other's personal identities is not going to be very effective! But it's still worth pointing out the two very different ways the term can be used, and important not to accidentally substitute one definition for the other.
That's somewhat in tension with being able to express dissent, in that being able to dissent constructively requires both social skill and an underlying environment of trust.
It's easy to cultivate only one of those (eg most OS mailing lists choose dissent) but cultivating both is harder and I haven't seen it ever happen by accident - all the communities that come to mind are ones that direct a lot of attention and effort to community structure and expectations.
I feel the "social rules" are quite valuable both inside and outside the workplace. I wish more employers and groups practiced them: https://www.recurse.com/manual#sec-environment
It's interesting how you choose to represent the safe space culture by its most annoying examples, but chose to represent tech company management culture by it's most positive examples.
However, most people let there identity and there ideas get tied together. So, it can feel similar.
And it is the responsibility of the person on the "attack" to make that clear using whatever approach works in the situation. Otherwise you have a communication problem, and it is primarily your fault for being inarticulate.
In any case, I went out of my way to not be snippy! I just thought it might be helpful to you to know that the mis-use slightly tripped up at least one person. It's hard to know who will find it helpful to have things like that pointed out (like some non-native speakers trying to learn) and who will just get annoyed. Mis-judged this one!
Also, they really are pronounced the same, if your hearing a difference it really is just in your head.
I don't think that's a fair characterisation of what "safe" in "safe space" means.
Is that not a safe space?
Contrariwise, the collegiate-style safe space makes it so that the certain people have boundless rights to be offended, but no corresponding responsibilities to consider other's feelings. In a team context this leads straight back to exactly the failing the article is primarily about, one person/group speaking a great deal more than the other.
(And please do not insult my intellectual integrity by claiming that collegiate-style safe spaces are supposed to be safe spaces for everybody. That is transparently false, to the point that you are just discrediting yourself if you try to claim otherwise. It is abundantly obvious who they are for and who they are aggressively not for.)
This makes me wonder if your only exposure to the concept is through the media, which tends to complain about students a lot.
Or I can feel unsafe even if technically we are perfectly aligned but I know they are going to backstab me in front of my boss anytime they feel like.
I can work or even share room with an Atheist, a Buddhist and a another Christian for weeks on end and we didn't bother another.
I think it is usually enough to respect another, and that includes for the easily offended part (often me) to respect others right to don't care about my customes as long as they don't mess with me.
I do however feel a tiny bit unsafe (for the lack of a better word) in online foras whenever I see people getting online-lynched for not having the correct opinion.
The problem is in attempting to attain the things above, I've also been on teams in a company where I was clearly the strongest person on the team, and was chastised for providing too much guidance because management had the idea it was better for people to fail on their own (in production) than it was for only one person on the team to be the primary contributor of plans/the way to go. Maybe they were right, but the end result was that it made /me/ feel disrespected by management, and my own contributions suffered, causing the overall team to suffer because the other team members were too junior to pick up the slack when my contributions fell off.
I think there's a right way and a wrong way to try to achieve these qualities in a team. My personal experience is that when you have a team of people who are all very strong in at least one skill set and have clear pathways to communication it works much better than having a team which is relatively weak/junior with only one or two people who are strong unless there are clear lines of seniority. It's a consequence of the idea in modern society that "all opinions are created equal" is not only true, but a good thing. It's important that either you're in a group where you can trust the equality of everyone's opinions, or that you have clear leaders that can mentor everyone else and provide strong guidance.
Basically, if you can, build an "A-Team". If you can't do that, make sure you don't actively undermine your A players. If your A players aren't jerks, they're naturally rise into leadership roles in the team and help mentor/teach everyone else to bring them up as well. Undermining the A players will end up causing the entire group to suffer and eventually the A players will leave and you'll be stuck with a group that's significantly weaker for it.
Of course my thoughts above are primarily focused on technical teams, I don't know if they apply as clearly to non-technical teams, and this Google study appears to have looked at both.
Like him I was also in a military setting before I became a programmer. I haven't yet experienced the same level of engagement and team cohesion outside of the military but I believe that the answers are to be found in the kinds of insights that studies like this uncover.
[1] http://www.tablegroup.com/books/dysfunctions
edit formatting
However, I have to chime in: I don't agree with the accountability aspect. Responsibility is necessary, along with knowledge and support from the whole system: this broad "ability to be effective" makes accountability obsolete, and is superior.
That combination of Knowledge, Psychology, Systems thinking, and a fourth: an understanding of variation and statistical reality -- leads to a much more complete and functional model of the organization. This is W. Edwards Deming's System of Profound Knowledge.
However, you can see why "The 5 Dysfunctions of a Team" caught on, and his name didn't.
Think of it this way: It's about the responsibility of members of this site to downvote non-productive comments. It's not about meeting a quota of productive comments; it's about disarming non-productive action. We hold each other accountable for having productive discussion.
More subtly, I'm saying that focusing more on accountability than on the conditions that enable fulfilling responsibilities creates negative feedback loops from the inherently negative reinforcement of accountability.
As the others have noted, accountability only exists as a means to control through fear. It only comes up in the event of failure, and is a means of personal individual punishment.
For that reason, it is not conductive to improvement, nor does it effectively motivate people to actually do work better over the long term. Invariably the limiting factor that caused them to fail in the first place is not a simple lack of care or willpower: the only factors that accountability is truly suited to control. Instead, failure is most often due to a complex web of factors both internal and external to the individual—and the only way to improve is to focus on that system. Accountability instead focuses all efforts inward and tends toward ignorance of the system, which is ignorance of the truth, which is a path to more failure.
In short, accountability is an unproductive concept born out of distrustful false beliefs about people. Discard it from your repertoire.
I believe it is possible, albeit in a slightly different form: elite sports teams and groups of people tied together with a strong shared purpose can get there I think.
edit and yes, as marktangotango says - shared hardship and knowing you can count on the next guy because he's been through what you have and hasn't broken either is also key.
So they took wastly (40% worse) suboptimal method of groupwork [1], which emphasizes only communication abilities and then they try to infer general claims from that.
Of course they will find that "emotional senitivity" makes "teams" that work in such conditions better.
And of course they won't find "common patterns for successful teams" in their meetings because, basically, meetings are not where decisions are made.
What a clueless research.
[1] http://observer.com/2015/05/brainstorming-does-not-work/
Would not you agree?
Paper says "Finally, c was positively and significantly correlated with the proportion of females in the group (r = 0.23, P = 0.007). However, this result appears to be largely mediated by social sensitivity (Sobelz = 1.93, P = 0.03), because (consistent with previous research) women in our sample scored better on the social sensitivity measure than men [t(441) = 3.42, P = 0.001]. In a regression analysis with the groups for which all three variables (social sensitivity, speaking turn variance, and percent female) were available, all had similar predictive power for c, although only social sensitivity reached statistical significance (b = 0.33, P = 0.05)."
[0] https://www.cs.cmu.edu/~ab/Salon/research/Woolley_et_al_Scie...
Some conclusions:
- Teams and focusing on team cohesion, team building, and team performance is far more important than individual performance.
- Understanding psychology (of individuals, of individuals in teams, and of organizations) is critical for making companies work. "Psych Safety" amounts to the foundation of Maslow's hierarchy of needs, as but one example.
This stuff is important. Build good teams, don't just hire good individuals.
And yet - most of us measure ourselves, how we are performing in a team by things that happen in team meetings(according to the article anyway):
1. Is idea my ideas heard and debated before being accepted/rejected?
2. Do I get opportunity to present my idea at all.
A time boxed meeting does not by nature leave room for idle chit-chat or from my experience even opportunity to hear everyone's thoughts. Part of the problem of course is, meeting rooms in lot of companies are a scarce resource, so you better clear out after time is out.
This is really curious because we hate meetings and yet we have adopted them in a sort of haphazard way which creates more problems than solves.
I meant 20-30 minute meetings that teams typically have, like retrospective or something.
But I was not talking about Agile per-se. I meant in general most meetings of software teams are time boxed nowadays- no matter which development methodology they are following. I am also not criticizing time boxed meetings since I have no data to back that claim.
But I am just curious, what if as a rule - teams decided to have fewer meetings but meeting last as long as they have to not because someone is knocking on the door for next slot. Will it make things easier?
Also one of the core ideas of Agile is adapting to suit your team, rather than following some process from a book as if it were written in stone. If time limits on certain meetings are causing your team more grief than they help, remove the timeboxes! Talk to your team and see what they think, that's what the retrospective is meant for.
Problematic stab at engineers. I don't think people go into software because they prefer computers to humans.
I've since improved my social skills a ton, but it's natural to expect different psychological trends in a population of engineers than in, say, people who go into sales.
The key takeaways that you find in very effective SOF teams are:
1. Clear and accepted authority chain (Top guy earned their place at the top)
2. Candid and constant communication/feedback between all members
3. A clearly defined, proven effective and specialized role for every member with each member knowing how each others role plays together
4. Common sense of purpose (Objective etc...)
5. Common experience (BUD/s, Q, FTC etc...)
If you have all of those then pretty much everything else falls into place. Note also that this has no requirement for everyone to being an "A" personality or whatever and has basically no bearing on what they do outside of work. Those aspects can help, but they can equally hurt depending on the egos involved.
https://hbr.org/2015/08/what-companies-can-learn-from-milita...
Team of Teams from McChrystal is a great book.
SOF: Special Operations Forces
BUD/s: Basic Underwater Demolition/SEAL Training Course
Q: Special Forces Training Course
FTC: Field Tradecraft Course
> One engineer, for instance, had told researchers that his team leader was "direct and straightforward, which creates a [psychological] safe space for you to take risks."
However the rest of the article went onto to talking more about expressing one's feelings and vulnerability.
I think this is a key difference between Google's and the US Military's approach to team building.
the military and google are very different organizations with different goals, so it's unlikely they will be run the same
a perfect team in the military will not look like a perfect team at google.
both organizations will wisely punt to other organizations when they come up against a challenge outside of their domains of expertise.
for example: the military does not design and build its own aircraft/weapons/tanks, nor does google engineer and build its datacenters
Ha, I actually did, but thanks. Also good guess on my team name (commando) ;)
You are right though that it's not a 1:1 substitution but the principles align well.
- pre-selection of qualified individuals
- dictatorial authority precluding much superfluous decision making
- martial and capital punishment casting a long shadow
- high stakes besides all of this
- weeding out according to fitness long before harmony is reached, where the weak are pruned mercilessly
- shitty HR negotiations
- lies during interviews, that can neither be proven nor disproved
- grooming habits
- sexual tension
- and the rest of all of our human foiblesI can't quite get my mind to really understand why some viewpoints respond with aggression to ideas like what the article calls psychological safety.
I think it might have to do with how it's the aggression itself that were being called on to remove.
These are terrible qualities for teams who are self-assigned, entrepreneurial, and have to disrupt standard accepted behaviors in order to create new products.
For example in Number 4: How does this work when the teams assignment is to come up with an objective, rather than just have one assigned to them.
Answer: They would splinter and struggle because they are following a chain of command.
Which follows to Number 1: To make products you need a diverse set of leaders who share authority in varying decisions, not a chain of people who wrongly think themselves to be universal experts. This does require Number 3, but in a different cultural fashion that isn't so rigidly hierarchical.
Basically these are great for teams who just take commands from the actual guy makes the decisions. Google and the NYT are interested in the teams that make decisions. I think the modern workplace is actually trying to UNDO these very narrow minded and ultimately damaging behaviors and their misapplication to the creation process.
What? Modern Western armed forces, and especially special forces units, make many independent decisions requiring great amounts of creativity and initiative. The hallmark of over 300 years of Western military units is the ability to operate without consulting the chain of command for every decision.
Were you a member of an armed force from a more centralized military, such as Russia or China? If so, then your perspective makes more sense.
There is a middle ground between Google and, say, Lockheed where teams are given a general direction to pull to keep them organized and working towards the company's goals but are otherwise free to determine how to accomplish those goals on their own.
SOF on the other hand are typically given a problem and asked to figure out a solution. Watch Charlie Wilson's war or read any of the books about the weeks after 9/11 in Afghanistan as a good example of this in practice.
For example in Number 4: How does this work when the teams assignment is to come up with an objective, rather than just have one assigned to them.
Answer: They would splinter and struggle because they are following a chain of command.
This is completely wrong. In fact the majority of the problems SOF are given are within the domain of objective identification for effects. When problems come down, everyone works together equally (back to the communication piece) to come up with a solution just like we do with startups. When it's time to execute, you execute within that chain and take direction based on your role.
Basically these are great for teams who just take commands from the actual guy makes the decisions.
Again, that's not how it works. Often the "commands" are even more nebulous than within narrow software domains. For example: "Keep [wartorn city] stable enough to allow for elections to run."
The "higher pay grade" problem a team must solve in your example is to shift a popular mindset in a region to inspire them to create elections, not keep the city safe while elections occur as assigned from your superior officer.
These teams need to decide the human social agenda, not just the strategy and execution of one given to them.
I think the difference is that you assume that is not happening in the entrepreneurial world. In fact it is, only guided by consumer demand, or some broadly defined corporate niche or group of uncoordinated individuals [1].
But that gets off the point however. Lets be clear here though, the scope of the article was how Google builds teams within the already defined "higher pay grade" problems. So by definition that is the context of what I am discussing teams already within a system. As it relates to Google they have "higher pay grade orders" from the Executive team of Page/Brin etc.... This is not Valve we are talking about here.
I think you are trying to make a larger statement about entreprenurial endeavors in sum total at the economy level which is not the scope of this discussion and it not relevant.
[1]https://en.wikipedia.org/wiki/The_Tyranny_of_Structurelessne...
We become dystopian when we apply the ultimately violent tactics of the military to the creation based goals of civilians.
Taking a simplified example to make further arguments is not fair. There can always be further counter-arguments. SOF will be deployed and then they adapt to problems with a clear goal.
If you told the team, develop plans for the stabilization of a government, they can do it.
I'm sorry, but you could not be more wrong. Even in the "conventional" Marine Corps infantry, we were not trained to ignore any problem. We were constantly being taught to take over the job of our superiors at a moment's notice. All downtime was filled with adhoc classes to train the lowest grunt to operate at several levels above his pay grade.
If you think SOF or even Marines defer silently to officers, you only know what you have seen on TV. You are just completely off base here.
At the bottom entry level units do defer. Boots stick around for morning formation after weekend leave still buzzed drunk because they're just told what to do and face consequences otherwise.
But then in the Corps the TIG promotions tend to weed those types out so that the motivated and qualified ones go upward the chain as NCOs. Then TIS promotions tend to be push those more qualified into SNCO roles.
To say that every single boot, or that every single service member is immediately capable of stepping up to be field officers because the one charge that was saluted all the time got sniped off is a ridiculous assertion.
But it is heavily incorporated, and more specifically in the Corps, for lateral movement up, or even down, a few ranks.
That's setup to optimize for the resiliency of the organization whereas in a corporation that's fragile because people working there are merely linked by bank account and therefore inspire no loyalty.
Massive difference.
Hence, I think it's massively important to realize that hiring for elite anything is really nothing like hiring for random warm bodies-at-problem techniques oftentimes used by mediocre performing companies that have a lack of leadership capabilities.
But there's still plenty of work for everyone else. Turning a 17 year old high school dropout into a team player with good attitude and good work ethic and good mind set is what this thread deviated to.
Any professional endeavor fundamentally is about scoping around a particular solution that will deliver value to your stakeholders.
They are specially selected and are non-representative. Whatever lessons that can be drawn there are inapplicable to conventional forces, let alone regular teams in the corporate environment. They might apply to special project teams ('skunk-works' type projects), but that has limited value as it is not scalable.
I imagine a large part of what SOF do is solve unforeseeable problems in the context of trying to accomplish some goal, and to be able to adapt quickly if the goal changes. This is pretty analogous to what a startup is faced with when trying to accomplish a goal, when you'll find that most of your team's initial ideas, as good as they seemed at the outset, were wrong, and you'll be trying to face problems and seize opportunities that you didn't even know you had. For both SOF and these teams, being able to adapt quickly to new information and circumstances, and to resolve the problems that come up which will destroy you, are critical.
I apply "startup" also in the sense of "startup-like" teams in large companies who are required to innovate rapidly.
I think we are in agreement - the applicability is likely to limited to 'small' specially chosen/specially tasked teams. i.e. it does not scale to thousands of employees (otherwise the army would be doing the same for conventional forces)
As a counterpoint everybody follows a chain of command. In all the corporate jobs I have held I never experienced a chain of command noticeably different from the military. Just like in corporate world the military allows you some latitude to push back on your leaders if they make a completely disastrous decision. The primary difference there is that you are more motivated to push back if a bad decision could mean increased risk to security (people's lives). I have never experienced this level of critical impasse in the corporate world.
The military is mostly like working for a corporate employer with some key differences:
* In theory the military expects everybody to be a leader, though in practice not everybody is willing to step up and make leader-like decisions or in some cases toxic leaders will suppress the opportunity.
* The military is a really big bureaucracy, which can constrict many creative (unorthodox) leadership decisions, but sometimes produces extremely unorthodox solutions to work around the bureaucracy.
* There are points in the military where you are demanded to work really long hours (like 12-16 hours, 6 or 7 days a week). In that amount of time everybody's personality is hyper-amplified. Some people can make this work really make their teams gel, where other people become terminal destructive forces.
* The military is often really bad at a constant work pace. Consider the phrase "hurry up and wait". You tend to get really good at accomplishing tasks at 4x human speed so that you can go back to being paid to self-study or watch movies.
* Look at how hard your corporate CEO works, the number of simultaneous tasks they have to balance, and the constant uncertainty in their schedules and travel. When you deploy in the military the entire team works at that tempo all the time.
* You don't get to be a conformist tool as a military technician, because then the bureaucracy will crush your soul. You quickly learn to invent your own solution to many common problems. This is substantially less true of many corporate software developers.
Works fine in case of actual universal experts Steve Jobs's, Jonathan Ive's, Larry Wall's or Linus Torvald's of the world.
That argument begins to break down when you appoint the wrong people to that job. You now have to follow orders only because some one is giving them, not because they are right.
Was it better to let everyone speak as much as they wanted, or should strong leaders end meandering debates? Was it more effective for people to openly disagree with one another, or should conflicts be played down? The data didn’t offer clear verdicts. In fact, the data sometimes pointed in opposite directions. The only thing worse than not finding a pattern is finding too many of them.
Once all of the debate is over, a decision needs to be made and followed through with, with the understanding that new information will require tweaks or wholesale re-evaluation of the plan until completion.
So in that sense, the leader ensures that everyone is executing their role well, and that everyone is aware of any deviations to it - while also delegating authority or decision making as needed as the task unfolds.
It's way more nuanced and delicate than it would seem from the outside.
I would hazard a guess the average Marine officer knows more about Kurt Gódel's incompleteness and its real life applications than the average techie.
They do for the kinds of teams I am discussing.
The most important thing I learned there was a Team Dynamics study, reportedly used by the DoD, which was distilled into four words: Forming Storming Norming Performing
What blew me away was that I could literally see this happening right in front of my eyes with a small team of kids who I was coaching for an FIRST Tech Challenge (FTC) Robotics competition.
I finally understood the team dynamics and realized that I am not the only one experiencing it and it was not an anomaly.
https://en.wikipedia.org/wiki/Tuckman%27s_stages_of_group_de...
http://www.dau.mil/pubscats/pubscats/pm/articles03/pat-ma03....
Oh boy, that's a huge piece of teambuilding training throughout the military. It's a good description of how new teams form. Unfortunately most teams don't get past the storming phase.
The other two are good with hands-off approaches. Heck, Performing can even take a certain amount of ineptness in leadership without falling apart.
Accenture/Facebook/McKinsey/Goldman all have common training programs that they put new hires through. It might vary by function, but almost everyone does it.
One caveat is that (in my limited experience) military hires tend to obsess a little too much on chain of command. Chain of command is important on the battlefield, but in corporate environments the best ideas come from anywhere. And in the best companies, people focus more on being "The best" than being "The leader".
I think the difference is that is important to have a clear process to make decisions. It doesn't need to be hierarchical, but it should avoid analysis paralysis or other ills of uncertainty.
1 - You decide without my input
2 - You decide with my input
3 - Joint decision equal input
4 - I decide with your input
5 - I decide without your input
Most good decisions are Type 2 or 4, though sometimes they need to be others. (Compliance, etc.) Then one can say, "This is a type 4 decision - I need to hear you out, but it's on me."
Many good ideas in the military and on the battlefield come from the ranks, too. The chain of command is about decision-making, as jacobr1 notes.
> And in the best companies, people focus more on being "The best" than being "The leader".
This is also true in the military, since there are dedicated tracks for enlisted and officers. But senior enlisted are still expected to lead (without being the senior officer), just as a senior employee should be expected to lead (without being the CEO).
Article about the study in English: http://sciencenordic.com/tough-special-forces-don’t-strut-th...
As for the sibling comment in this thread wrt: "breaking down and building up tweens" -- while the training to get into the Norwegian Special Forces training is though, and the subjects are often young -- I think there might be more of a cross-over in organization/leadership/group dynamics than what might appear on the surface.
There are a couple of things that are needed for military operations that are different from similar activities (eg: search and rescue) -- and parts of the military mind does go against what could be considered a sound human mind in today's society. War is inherently about violence and murder. But even if we do not want our programmers to actually be assassins/ninjas -- don't want them to actually have a killer mindset -- much of military training is about changing one's idea of what can be achieved by an individual and a team (to realize that one can indeed do the (previously) "impossible". To not give up, and part of the intense drills are part of making someone an expert. In physical arts, intense drilling is required for technique to become "good enough" for expert level (this is also AFAIK true for dancing, for example).
One should also remember that as long as the organization in question is able to do meaningful self-reflection and improvement, training methods can change and result improve.
Excerpt:
As the researchers studied the groups, however, they noticed two behaviors that all the good teams generally shared. First, on the good teams, members spoke in roughly the same proportion, a phenomenon the researchers referred to as "equality in distribution of conversational turn-taking." On some teams, everyone spoke during each task; on others, leadership shifted among teammates from assignment to assignment. But in each case, by the end of the day, everyone had spoken roughly the same amount. "As long as everyone got a chance to talk, the team did well," Woolley said. "But if only one person or a small group spoke all the time, the collective intelligence declined."
Second, the good teams all had high "average social sensitivity" — a fancy way of saying they were skilled at intuiting how others felt...
That is way more fluid in terms of authority and touchy-feely than rigid military hierarchies.
Which makes sense given that roles are much more fluid in creative enterprises and there's no life-or-death quick decisionmaking that favors decisiveness over debate. Not to say that there's no creative thinking in the military, but role and responsibilities can be more clearly defined at the outset.
CONOPS (Concept of operations) planning goes exactly like described with equality of input throughout the process. In fact it goes beyond that with team members continuously providing input throughout the run-up and execution of whatever operation is in play.
In many cases during whatever operation is happening, any team member, regardless of rank, can call "knock it off" if something is not going well and the team will respond to that.
That is way more fluid in terms of authority and touchy-feely than rigid military hierarchies
Again, this rigidity of hierarchy is a myth within high performing SOF teams (CAG (Delta), DEVGRU, 75th Ranger Regiment etc...)
there's no life-or-death quick decisionmaking that favors decisiveness over debate.
Again, this misrepresents how battlefield decisions get made and overplays how democratic decisions are within [insert startup]. There is advantage to "violence of action" - meaning making a decision can be made and emphatically executed quickly - but it's not the case that it needs to be approved by TOP (the leader). One of the key benefits of SOF is that everyone has the ability to make a decision that might veer from the plan without it being insubordination or some other offense.
people may speak over one another, go on tangents and socialize instead of remaining focused on the agenda. The team may seem inefficient to a casual observer. But all the team members speak as much as they need to. They are sensitive to one another’s moods and share personal stories and emotions.
Overwhelmingly the key point of the article is "In the best teams, members listen to one another and show sensitivity to feelings and needs." It focuses heavily on emotional and psychological needs, not "clear authority" or "clearly defined, proven effective and specialized roles," which is what the comment was trying to hijack this to emphasize (that stuff is nowhere in the article).
It's true the previous comment did include a line on "candid and constant communication," but that is quite overbroad and omits the key distinction of the nature of that communication. The key point of the article is that it is of a tangential and/or personal nature, not just candid operational feedback. This is why I think it's fair to criticize the comment as missing the whole point of the article.
This is insightful, though comparing SOF teams and commercial work teams fails in the critical areas of complex interdependency and trust. [0],[1]
I think what this google study highlights most is, candidates are not selected for man management and emotional robustness: ie: emotional honesty, self awareness and the ability to equally lead and follow. This cannot be taught. Emotions are traits. They can be shaped somewhat through team assignment as shown with the experience of Matt Sakaguchi.
[0] Barbara D. Adams, Ph.D. and Robert D.G. Webb, Ph.D., "Trust in Small Military Teams" A great read on the qualities of complex interdependent teams ~ http://www.dodccrp.org/events/7th_ICCRTS/Tracks/pdf/006.PDF
[1] You fail, people around you get dead. Those around you fail, visa versa. The mission is compromised. You do not see sacrifice & self sacrifice like this in commercial world.
At the very least - [citation needed]. The medical community seems to disagree on that when it comes to teaching self awareness, at least for the majority of students[1][2].
I'm not sure what you mean by "emotional honesty", so it's hard to reply to that. If that's the equivalent of emotional intelligence - yes, that is teachable. Quite well.
[1] http://www.bpsmedicine.msu.edu/pdf/22-%20%20ctstudypstrng.pd... [2] http://www.aacp.org/resources/education/cape/Documents/Habit...
Emotional honesty is pretty simple. Will you speak up when necessary, even if it contradicts peers or the leader? What was the last time you told the boss, "this is a stupid idea Boss, change it, or this will happen" and then the boss does so without recrimination? Remember the outcome, not ego matters.
As for the field of medicine, of which I know nothing about, in Melbourne (Aus), selection into health sciences is filtered by both self awareness, emotional maturity and smarts via both academic results, independent Uni examination (Monash) and panel interview. This the "spectrum" filter for those sufficiently smart enough to get through the academics, yet are more suited to research.
Could you teach MDs to program and work at google? Probably. Could you teach the clinical skills MDs require to make medical decisions on patients to Google SEs? Probably not.
Doctors work with people. Google engineers work with languages and data on silicon.
But that aside - your reply has nothing to do with what the studies I shared demonstrated: Emotional awareness is teachable. Your idea that it isn't seems entirely based on your personal assumptions. (If you have actual evidence, please share)
As for "emotional honesty" - that's pretty much the same as honesty, no? There's no emotional component to it.
This is much easier for a SEAL team, which has "someone else" dealing with all the other stuff, from procurement to catering to writing the paychecks. As a result, the same rule will work for making teams in an enterprise, but a SMB cannot do this - there isn't much room for "not in my job description" at a SMB.
90% of my current job is doing what is nobody's job. (10% is my 'main area of responsibility' :)
https://en.wikipedia.org/wiki/Kunduz_hospital_airstrike
may, in just 45 minutes, very well have undone all the putative good the U.S. and its allies were trying to accomplish in all of their 14 years in Afghanistan.
So on balance it seems that the culture of elite U.S. military unites is something to be shunned and avoided in civilian contexts, rather than followed and replicated.
I started cracking up at this point because me and every engineer I know would have said the same thing in response to a survey designed to promote team building.
When people mostly pick who they work with and how, it's usually quite automatic and natural for things to work out without managerial baby-sitting.
And I would completely agree with that. In STEM I'd say the majority does not understand the concept.
I'm also not saying I agree with this mindset. I'm just answering your question: > What is it with Silicon Valley that always ends up in "you spend all your life at work and it's a good thing" ?
If your standing within a group is assured, you can call each other dumb assholes all day and it won't matter. This is the classic phenomenon of male friendship. What works in a group isn't an absence of criticism, it's a sensation of security of your belonging in that group.
How do you make people feel secure? The only real way it occurs is if everyone actually has respect for one another's contributions and work. This can't be faked, unfortunately, and respect has to be earned, and everyone's respect is earned slightly differently. Some peoples' respect algorithms are more compatible than others.
An interesting study might be to try organizing people by what they value most in their co-workers.
Why do you single this out as an especially male phenomenon?
These results seem like a condemnation of any management practice based on competitive exclusion.
If anything, I've seen the lack of sharing a bigger issue in places with a culture of competing metrics. "You get paid on revenue, and I get paid on time to completion." I've seen terrible sharing behaviors absent the forced ranking.
Now that I've felt the hammer of that first hand I want nothing to do with it. I try and be conscious of my teammates. What are they feeing? Do they have an idea that they are scared to share? What are they scared of?
Does anybody have an idea what that survey was or looked like?
If you think you are very good at this and don't need conclusions driven from poorly controlled experiments, please do yourself (and others) a favor and read through!
So, over the next 10 years and many products, I'm extremely proud to say, none of those products ended up failing. And, most eventually exceeded expectations. Who knows, it could have been by chance. Or, maybe it was regression toward the mean. But, I like to think that my focus on the problem and my motivation helped solve it.
Over all of this, I was rarely directly credited, which is not surprising. Often I wasn't the 'manager' or even the 'technical lead'. I was just some relatively smart dude on the team, and I was in it for the thrill and enjoyment.
The key is to
1) be a good manipulator (or, more politically correctly, a good 'influencer')
2) clearly understand the strengths, personalities, motivations, and goals of each member of the team
3) view each person as a potential 'force multiplier' (to take another word from the military)
4) strategically coordinate the right activities between the right people at the right time
5) lead from the bottom, by example
6) be positive and optimistic
7) recognize potential conflict early and extinguish it subtly
If you have a really awesome manager, they can do this from the start. But, as many of us know, management (in the U.S., at least) is a status position, and it attracts those who like status or competition. I've only worked with one manager (out of dozens) who has been able to do this, well. It's a rare skill.
Like him I was also in a military setting before I became a programmer. I haven't yet experienced the same level of engagement and team cohesion outside of the military but I believe that the answers are to be found in the kinds of insights that studies like this uncover.
Differentiation comes from "feeling psychologically safe" so you are able to be creative and express your ideas. When you feel safe to be different you do your most creative work. In the beginning of the article I thought the author will conclude that teams made of people with different skill sets would work better than teams with more similar members. The Google 20% time for creative work is also boosting differentiation through creativity.
Integration comes from balancing speaking and listening. When everyone can speak, they become mutually integrated. Your ideas have a fair chance of being heard and influencing the group decisions.
The same pair of principles (differentiation and integration) are present in other systems like the brain, the ecosystem, the free market and the open source community. I originally discovered these concepts from Giulio Tononi's consciousness theory of integrated information. The same two ideas interestingly coincide in name with the basic operations of calculus and are founding principles in the European Union.
There is nothing wrong with this article and their methodology because it's not about real science at all. Science produces models, with predictability (among other things) where you or I could go out and get the same results. That is far from what Google has actually demonstrated - and I don't know whether they intended to do that in the first place.
Psychology has never been accepted as a science because they don't produce accurate, reproducible models. I don't want to sound rude, but all this stuff isn't exactly new and can be searched for online. A random set of links from 10 seconds of googling:
https://www.theguardian.com/science/2015/aug/27/study-delive...
https://www.newscientist.com/article/dn20447-journal-rejects...
http://psychsciencenotes.blogspot.com/2011/05/failing-to-rep...
What would you call this thing in the article that people did then? Statistical analysis?
Now this notion of psychological safety is interesting because it makes me think of our larger environment (too much a leap!) that it's similar to our personal safety ( perhaps requiring more surveillance?). It reminds me of Hayek’s notion in "Road to serfdom" opining that submitting to less freedom for more safety will make you worse off. I do have to say I would probably submit to more safety - it's so much easier.
In modern animal breeding, our goal is typically to improve some trait through genetic means. We extract the animal's DNA, sequence it, and compare it across many different families in order to evaluate how that individual differs from others and how its genetics affects our trait of interest. We call this trait our 'phenotype', and in animal breeding, phenotype is king. We can do all of the analysis we want, but if our phenotype is inaccurately measured or uninformative, it's all a waste of time.
Has NYT ever done an article in this format before?
The "Blind Hiring" section starts off great:
A few years ago, Kedar Iyer, an entrepreneur in Silicon Valley, became acutely aware of a problem in his industry: A surfeit of talented coders were routinely overlooked by employers because they lacked elite pedigrees. Hiring managers, he thought, were too often swayed by the name of a fancy college on a résumé.
It's an excerpt from a book:
> Charles Duhigg is a Pulitzer Prize-winning reporter for The New York Times and the paper’s senior editor of live journalism. He is the author of ‘‘The Power of Habit’’ and the forthcoming book ‘‘Smarter Faster Better: The Secrets of Productivity in Life and Business,’’ from which this article is adapted.
Is this what is required to have a team jell at work? I imagine the same sentiment could have been achieved with a different question, but the article continually references this one. Does it make me a non team-player if I don't want to share my health issues with my coworkers? (In addition: is this legal to do?)
I wonder if giving this test during interviews would help in building better teams.
https://books.google.ca/books/about/The_Affinities.html?id=z...
This is of course 100% anecdotal.
This of course runs 180 degrees against the concept of fungible engineering and Google's generalist fetish.
Or TLDR: I prefer building the A-Team instead of trying to be Agile.
There is fundamentally too little time in the day, and too much to know to truly be GREAT in a domain, to expect anyone outside a select few (certainly not enough to fill out general employment at google) to be able to bridge these domains meaningfully and still be as deep as possible.
Now; there is a massive benefit to the specialists having sufficient understanding to communicate with other domains. But this to me is not what being a generalist means, and is an asset that is often looked over even in the generalist-focused agile orgs I've worked for; "are we communicating well internally in general".
(To bring this back to the article as well, I've also found the specialist groups with "eigenvectors" of skills to be MORE cohesive, as it's a great trust builder when someone can say "I'll be supporting your work in Y with X" and you know that they'll be doing X _light years_ better than you could, and you can truly TRUST them to have your back. Not that the same thing isn't very feasible in other environments, but that's an "easy way in" to trust.)
One-man bands are a different story though, if a full engineering team is not feasible, a generalist who has a broad but shallow knowledge of lots of things is far preferable to a specialist of one.
In the outside world, the success of a company can be measured quite objectively; but inside one company, what defines success?
Isn't it possible that patterns are hard to find because "success" is vague and, fundamentally, random?
How do you get a group of cynical alphas to take-off their game-face for work? This screams for a How To manual.
In other words, a very worthy read.
The article in this post echos what I had felt: https://news.ycombinator.com/item?id=10731071
This is an interesting article since he cites a pub quiz and sports - something I participate in.
I've been playing for about 4 years in a pub quiz. We've won once and are considered a strong team. However, the difference between the a strong team, and the better team in a night, is the fact that our team fosters a "no suggestion is a dumb suggestion" policy. No matter how stupid you think your answer is, you are encouraged to say it. In fact, if we see body language from a person hesitating we're quick to jump in and say "Spit it out! Just say it!"
We essentially create a safe zone for our team members to not worry about the consequences of a wrong answer. There were so many times where a completely wrong suggestion makes another teammate say, "Hey! Wait, I think I remember!"
We also have a policy of if a person has a different answer than the rest, then he/she has to "fight for it". That is, you have to convince the rest of the team. By having a "fight for it" rule, we put controlled confrontation on the centre of the table and let people hash it out. There is no regret, or fear, or worrying about feelings. It's just a normal part of our evening that is done with humour and friendliness. Firm, but friendly!
By doing the above, we instil confidence in every team member. Those that are more confident by their nature, can still be challenged by anyone and keeps them in check, and the weaker confident ones feel safer to step forward when they need to. It balances out and our team has absolutely great chemistry because of it.
I've also played team sports for most of my life and some of the best teams I've played for had former professional athletes. Since they were stronger than the rest and more confident, they always raised their hand and took the blame for any mistake! It was quite funny because we knew my screw up was not their fault, but they would make an excuse about how they should've done 'X' and I wouldn't have screwed up.
Needless to say, that allowed players like me that weren't former pros, to be at ease and to be more confident and give my best, knowing that the strongest player on the team wasn't going to look down on me for every mistake.
It was a very interesting dynamic where the more confident person ensured that the less confident person is playing their best and it raised their confidence. Again, it is a kind of delicate balance that can change on any given night.
Do you mean Shockwave, or ads? ;)
Granted, I've also worked with some really great people, better than other places I've worked. And I'm sure unconsciously been a jerk myself.
So my summary is: run all the studies you want. People mostly behave in the personality patterns they set down as kids.
And most of us grew up in schoolyards full of not-so-nice dynamics.
And where is the discussion of sociopathy and other classic psychological manipulations? You'd figure that subtle disruptors would be critical to the study, since I argue that middle management collapses in most large organizations due to the Machiavellian predilections of petty power brokering.
If the teammates don't respect each other, they will ignore what the others are saying, even if they are forced to listen to them.