I’ve been called an “organizational expert” because of some of my writing, but I often have doubts about whether I really know what’s going on. Sure, I’ve learned a lot, but it’s mostly been through observing other peoples’ failures. I don’t get to see a lot of successes first-hand, because successes tend to end up being swarmed by people with more connections than I’ve got, but I’ve seen several lifetimes’ worth of failure and mismanagement. It’s not the career that I’d want, and while I’m aware of individual pitfalls and systemic weaknesses, I’m not sure that I’m one to give advice. I’m 32, and some of my peers are quant traders making half a million (or more) per year or getting seed rounds shoved down their throat, and my situation is… not that. I’m every bit as capable as they are; in some cases, visibly moreso. So what went wrong? What did I miss that they didn’t? I have a compact answer for that, and I think that it’s one that will be genuinely useful to my readers. Here it is: since leaving school, I’ve never (with one exception [1]) been on a winning team.
[1] What’s that exception? In 2008, I left a prop shop to join a startup. It was a mistake. Meanwhile, the quant/trading job market shat the bed, and 2009-10 was not a time to get back in.
I can imagine negative responses, from strangers as well as colleagues in the past, on that one. So let me say what I mean and what I don’t mean. I’m not trying to throw everyone I’ve worked with under the bus. I’ve been on good teams with some excellent people on them. I’d never say otherwise. I’ve been on high-talent teams that were well-run– but that lost, politically, within the larger organization. I’ve had good managers (although they tend to be replaced with bad ones over time, because skill at “managing up” correlates positively with being a bad human being). I’ve been on “star teams” that had more transfer inquiries than they could take, by orders of magnitude, but that inevitably failed politically and were unable to achieve much with the talent that they had.
As I’m getting older, I’m increasingly aware of the optical importance of being able to say that you were on a winning team. The projects that put a “Wow” factor into one’s CV usually come from teams, not individuals. Now, of course, one can lie about one’s history and career. Almost everyone does, to some extent. It’s still better to have the truth on your side. The low-grade, social-status-inflating lies that we use to bolster our professional images aren’t genuinely unethical, and we’re rarely caught out on specific ones, but there’s a cognitive load involved in maintaining them, and it’s better not to have to do the work. It’s similar to the “Harvard Effect”. I didn’t go to Harvard, and I don’t have a PhD, and I can convince a highly intelligent person that I’m her peer in 47 seconds. If I had 2-minute elevator ride with any investor on Sand Hill Road, I could leave him convinced that, yes, I’m better than most of the people he’s funded. The advantage of the Harvard PhD is that he can convince the same person of that in 0 seconds, and put 100% of his cognitive and social effort into being likable. Sometimes, that 47-second difference is meaningless, and sometimes it’s critical. It’s the same when it comes to upgrading your career. Most people have to do it, and lots of people do it subconsciously, and the people who are the most effective at it tend to be those who can discard what actually happened and create their own self-consistent truth in which they actually believe. Whatever works; I’m not here to judge, and I’ve seen so much shit that is actually harmfully unethical that, even though I don’t use them myself, literal CV lies (fudged titles, improved dates) don’t faze me. I do believe, even still, that there’s a not-inconsiderable social advantage if you’re in that sub-5% (1%?) contingent for whom the career that everyone tries to present is what you actually have. Again, that means that you can put 100% of your cognitive investment into being likable, instead of being distracted by having to maintain consistency while bending the truth.
What is it about winning teams that matters so much? First, individual accomplishments will almost always be downgraded, even to the point of being called trivial. Think of the 100 million people who plan on “doing a little writing someday” and the much smaller number of people who actually sit down and write novels. What’s the difference? Talent’s part of it (although it’s possible to write “a novel” and even publish it without much of that) and then there’s the huge amount of work involved in writing a good one. Fiction difficulty scales faster-than-linearly (a 50,000-word novel isn’t five times as much work as a 10,000-word story; it’s probably 10-20x) and anything readable requires revision at multiple levels… and then we have editing, publication, and marketing. Writing looks easy. A good finished product looks like it could have been put together by the reader. It’s not easy. The same can be said of designing games, computer programming, building technological systems, learning a new language in adulthood, becoming an expert in a technical field, and much more. Since this work tends to happen in private and without cinematic flair (there’s no Rocky montage; just lots of work at weird-ass hours while the rest of the world parties) there is a tendency of people to underestimate, massively, how much work it would take them to match an individual accomplishment. By definition, an individual accomplishment can be done by one person, and while that certainly doesn’t mean that it can be done by everyone, people rarely like to admit that someone else did something that they couldn’t. It’s easier to admit that you couldn’t build the Statue of Liberty (clearly, a group effort) than to admit that you couldn’t have written Hamlet, because a human out there actually did the latter.
There’s a more subtle point of resistance. Most corporate CEOs would probably admit that I’m a better programmer (or writer, or game designer) than they could ever be. That’s not a hit to their pride, because they don’t think that I could do their job. They don’t measure themselves on technical skills, because they’re too busy to try to become good at everything. Instead, they measure themselves by the quality of people they can recruit to work for them. What would be hard for them to admit (and what would lead them to down-grade my accomplishments) is that they probably couldn’t recruit and hire someone as good as I am. Am I really so daft as to say that? Well, yes, with some caveats. If a CEO created a $500,000-per-year programming position and marketed the hell out of it, he’d get hundreds, if not thousands, of inquiries. It’s a statistical certainty that he’d have at least one programmer better than I am in that pile. If he could hire all of them, then he’d have at least one person (and probably more) better than me (and quite a few bozos). In the real world, where he actually has to sort through the mix: could he pick the right ones, though? I highly doubt it. Of course, he might get lucky, but to have any reliability in that endeavor, he’d have to hire someone like me to sort through all of those CVs and spot the talent, so we essentially have him needing a dragon to catch a dragon. It’s rare that a CEO can admit to this, though. Most CEOs’ egos require them to look at any individual accomplishment and say, “with my resources, I could hire someone to do it better.” And sure, it’s possible to just get lucky, because the ability to recognize high-level talent is so rare that there are surprisingly many underutilized and incredibly talented people. (Give the female ones and the ones with more melanin and the ones from state schools a second look.) I’m sure that even the most dysfunctional large company has at least one employee smarter than I am. Can the executives find her, internally? I doubt it. Look at all the acqui-hires that go on: those are cases of companies purchasing mediocre talent at a panic price because their middle management filters have failed, and they’ve given up on their ability recognize the people like me (statistically, there ought to be at least a few) at the bottom.
Individual accomplishments, I’ve realized, will consistently be underrated by other people in adulthood. It’s the opposite of childhood, in which individual accomplishments are lauded. In adulthood, individual performance isn’t even measurable, because things worth doing (by an adult standard) are those that haven’t been done before. At any rate, people who are used to winning at easy things like climbing the corporate ladder (which isn’t easy for me, but is stupidly easy for well-connected people with underweight consciences) tend to think that they could just as easily do hard things (that novel, that startup, that “web game”) if they just had the time. Some of them possibly could, but it also doesn’t matter. (I could make the same claim about various billion-dollar companies that I “could have” founded.) Things are ultimately done by those who do them, not those who “could do” them. Even still, it’s hard to come up with an individual accomplishment that doesn’t leave a large number of people with some combination of “I could have done that if I just had X” or “I could hire someone off the street to do that.”
A larger problem with individual achievement is that it’s hard to make it into political success. One person going alone can almost never both achieve something meaningful and defend that achievement politically at the same time. You get stuck having to choose one or the other, which means that either you do great work that no one cares about (and may get fired) or you campaign successfully for something that you can’t actually deliver, have to scurry to build a team around it, and end up as a middle manager. Justifying work not only takes time and emotional energy away from work, but often destroys a person’s creativity and motivation. Realistically, any project will need some effort to go into the work itself, and some muscle to go into defending it, in order to keep the talent and people and resources from being paperclip-maximized by the rest of the company. Every company has these powerful warlords (sometimes official managers, and sometimes upstarts trying to put themselves in charge) who will harangue you, if you’re an individual high performer, over the appearance of favoring your own career goals over their career goals. Going out alone, you can’t defend against them. You will get paperclip-maximized, and you will start to hate your work and your company.
That’s more dangerous than it should be, because visible high performance often hurts a person, because people don’t face militant political adversity (i.e. risk of being fired) for being low performers. They’re most likely to get in trouble when there’s a visible change in performance. Long-term low performers get a pass, because their bosses are afraid that they might actually be working on something important that no one knows about, but middling players who drop to low performance, or high performers who drop into the middle, are seen as a present morale threat. When there’s a drop in someone’s performance, the managerial fear pertains to “the example” being set. If the office slacker drops to 6 hours in the office and 1.5 of actual work, the manager makes a note that he’ll be the first one to go if there’s ever a layoff. When the star drops to a 6-hour day, even if she accomplishes more than most people do in 12, the boss has to take action quickly, lest the team “catch” her motivational rhinovirus. Obviously, not everyone who takes a short-term dip in motivation or performance gets fired, but there is a much higher risk for that person than for the career low-performer (who’s probably become very good at hiding, having had years to perfect the art).
It’s easier for a team to defend itself, politically, than an individual. I tend to see this as a surface-area problem. If you double the radius of a sphere, its surface area quadruples, but its volume octuples. Its attack surface doesn’t grow as fast as its capability. Of course, there’s a counterpoint to be made (at least, in technology) about the sub-linear scaling of software efforts, and I think that that’s true at the code level. From a line-by-line source-code perspective, the optimal team size is 1. However, not only are there plenty of projects that require multiple people who don’t have to be trampling in each others’ code space, but defending a high-profile single-person project is almost impossible to do while also delivering the project. Even if the team of 10 is only four times as productive as an individual would be on the technical matters, it might grow faster than its attackable surface area. In the sphere example, said surface area grows with the 0.6667th power of its volume; in the corporate world, when we’re talking about a team’s political vulnerability, I think that the exponent might be much lower: possibly 0.2 or 0.3. Why so? Assessing that understands insight into how corporate politics works. I’m going to try to be charitable to business executives as a class– they’re not all scumbags, they’re not all stupid– and nonetheless assert that they’re cognitively exhausted. To be honest, I probably would be too, in their jobs. The privilege of being the CEO is that you can talk to anyone; the drawback is that everyone’s lying to you. The thing that pisses off an executive most (and, often, managers who are managing up into that class, in the hope of joining it) is “one more thing to think about”. A self-initiated skunk-works project, in this light, isn’t less costly than one with 30 people behind it. Office politics begins with executives “flipping the switch” on ideas and people when they’re tired of “one more thing to think about”, and trickles down into management who don’t want “one more thing to justify”. This should also explain why companies allow such inefficient large teams, even in the light of sub-linear programmer scaling, to exist. I’d guess that a typical team of 32 programmers is only 6-8 times as productive as an individual, but it’s only 2-4 times as attackable. Another way to look at this is that its per-person vulnerability has been reduced by a factor of 8-16 and that its capability-to-vulnerability ratio has improved 2- to 4-fold. Of course, even still, many teams fail to protect themselves against the paperclip-maximizing and cost-cutting impulses that are the id of the organization.
I’ve been a high-power individual (most of the time, in fact) and I’ve been on high-power teams. I’ve even been on “star” teams where everyone wanted to work, and whose political failure was not the team’s fault. In one case, it was a merger that went extremely poorly. Great manager, great people, great projects… and it didn’t matter in the context of a multi-company deal that was put together by scumbags and idiots, most of whom I’ve never met. That’s the “shit happens” factor.
This pattern, however, is a failure of mine. Make no mistake about it. I’m not passing the buck for my own career shortfalls on to other people. If I’ve had this many chances to pick a winning team, to not much success, well… The only common element in this pattern in me. I made a number of choices that were incorrect or even unwise. I consciously optimized for certain things, and they were the wrong things. I built up a strong skill set, becoming a high-power individual, which is nice, but now my capability/credibility distance is so severe that unless someone takes a huge chance on me, I’ll be three levels below what I’m capable of, and struggle with boredom. What did I tend to do wrong? Often, I chose jobs because they were at “good companies”, or because they “had good people”. That means nothing. A politically-losing team at a good company is going to hurt your career, and even dysfunctional companies are full of good people (and some bad ones, because it doesn’t take much of those) and it doesn’t make a bit of difference. Sometimes I picked “good teams” that weren’t politically successful. That was probably the most common mistake, certainly a more common occurrence than landing on a shitty team. Good managers aren’t (despite my complaints about the moral compromises that come with the job) that rare, but what can a good manager do if he has to fight to save his own skin? A bad manager in a good company is a disaster (ask me about Google, some time) but a good manager in a bad company can’t do much, and will typically get frustrated and leave.
Winning teams matter because individual accomplishments are so easily for others to armchair-quarterback and discount to “trivial” status. Even if the achievement required creative brilliance, presentation requires simplicity and begets a sense of “I could have done that.” The difference between me and other peers, at my age and talent level, who are making millions as quant traders or raising seed rounds for their third startup, is that they found winning teams early in their career. Was it luck, or skill on their part? I don’t know for sure, but I tend not to think that it was luck. It’d be more charitable to me to say that they got lucky and I didn’t, but I really think that that’s not true. I’d bet that this is a learnable skill that they had at 22 and that I’m just starting to get (and, quite possibly, too late).
I’ll note that winning teams are rare, because it is the way of corporate politics that most projects are shut down at some embarrassing, intermediate state. This is probably why one genuine winning-team experience bumps a person’s earning potential by at least 40 percent, while individual achievements tend to get overlooked. Good teams are not rare, or especially hard to find; winning teams are much harder to spot. In a job interview, it’s tough to suss this detail out. More to the point, I don’t think that I made it a priority when I was young. Good projects and good colleagues can make a job beneficial in the short term and, when I was younger, that was enough for me. If you’re willing to job hop every 12 to 18 months, you probably care more about the project and technology stack and general strength of colleagues than about the political strength (or lack thereof) of the team you’ll be joining. Over the years, though, that political armor is the only thing that prevents other powermongers (who, over time, might become your official managers) from paperclip-maximizing you into the pursuit of their own career objectives. Having done quite a bit of that job hopping, I’m tiring of it and I don’t like how it seems to end. So, of course, if I could do it all again, I’d pay more attention to the political status of each team within the larger company. That’s far more boring of a detail than neat programming languages or company prestige or “20 percent time” (which is a trap, as I explained here) but it’s also more powerful.
Realistically, most job interviews are going to be conducted by teams and managers in the company’s political mid-pack. Losing teams generally don’t have much hiring authority, and winning teams can fill their ranks internally and through referrals. That said, I can come up with one question that I think might have some signal. For a warning, this is an aggressive question and might cost you the offer. That question is: “How much time do you spend justifying the work of your reports?”
On a winning team, your boss will rarely, if ever, have to justify your work to anyone. Certainly, he’ll never be expected to account for you individually. (If you’re on a losing team, your second-level manager expects weekly reports on what you’re doing. That’s bad.) To be on a politically winning team means that, unless your boss dislikes you, you’ll have a lot of freedom in how you work. On a losing team (and most teams are losing teams) your boss will be held accountable for you and the pittance you draw. See, bad bosses can present themselves as good ones– that’s a risk that you’ll always take– but you might be able, through these sorts of questions about how work is assigned and justified, to get an honest signal about the manager’s relationship with the company and the team’s political armor. Of course, there’s a huge caveat. If you’re too picky on this sort of thing, you risk adverse selection, because you might land with the manager who tells the best story rather than the one who is most able to protect you.
What about joining a company and then transferring internally to a winning team? This is also dicey. I’ve never seen a company where everyone was an idiot or where every team was “bad” in the sense of having bad people, but I’ve seen many companies with zero winning teams (except for “the leadership team”, a gag-inducing euphemism for “executives”, which wins by having control over the rules). So there are many companies where there just aren’t any options. Of course, at some size, it becomes likely that there are at least a few winning teams. Google has some, as do Amazon and Microsoft. Can one join them, via internal transfer? At most companies, moving internally to a winning team is impossible, because so many people want to join them that, in a closed-allocation shop, the only people who can are those who are already politically successful. (If you think that corporations can tell the difference between “performance”/merit and political success, you’re on the wrong blog. This place is for adults.) People in closed-allocation companies are lucky to get lateral transfers, and those require not only a lot of work, but an insane amount of risk– because the first person to get fired when a team’s losingness results in headcount reductions is going to be the person who seems to be looking for a transfer.
In fact, I think the best way to get on to a winning team at a large company is to pursue what is, otherwise, almost always a career mistake: get (and take) a counteroffer. In general, taking a counteroffer is a bad move, because it means that you’re sticking in the same job (with a worse reputation, and shorter life expectancy) for an improvement in salary that’s either (a) too small to merit the increased risk of job loss or (b) large enough to signify that you were drastically underpaid, and ought to consider abandoning the relationship outright. Sometimes counteroffers include title improvements and management opportunities, but that means that one will be working at a company that offers authority in order prevent attrition (which, yes, is a common “anti-pattern” in technology). It’s almost always better to take a new job than go with the counteroffer that’s made in response to the new job. If the objective, however, is just to get on a winning team somewhere, then the counteroffer might have better odds– you know if the new job comes with a better title and salary, but not if it’s on a politically winning team. Of course, there are some difficulties with this approach. First, one typically has to propose the team change during the counteroffer negotiation. It won’t be offered, so it has to be part of a counter-counteroffer. (“I wanted $X, you offered $0.7X. I think I’d be actually be worth $1.5X to you on [star team], so how about we meet at $0.85X and I work for them?”) Second, counteroffers are usually only made when the company knows that a person is grossly undercompensated, which isn’t a state one wants to be in, in the first place. If it’s any other management failure (wrong team, wrong project) the company will usually let the person go sans counteroffer (at which point, one can’t stay in one’s original position.) Third, this play can only be made once per company, because two counteroffers is “a pattern” and once you have the reputation of being a flight risk, the company is going to do everything it can to keep you away from important work. So this is, all told, a strategy that puts a lot at risk and might not work. It is, however, one way to move to a politically successful team that probably has better odds than anything else.
What about, going further into the wild, alternatives to the traditional corporation? In true startups (less than 15 people) and small hedge funds, the importance of picking a winning team is even more important, but at least it’s more transparent. Here, winning team means winning company. I wish I could say that I knew how to predict winners and losers among upstart companies, but there’s so much noise in that game that I tend to doubt that anyone’s got it figured out. Even the venture capitalists who manipulate success-inducing factors like company reputation have a hard time making money, so what odds do I have? Just as good people can constitute a politically losing team in a company, or vice versa, the same seems to be true of startups. Evan Spiegel is not a high-quality person, and I struggle to imagine a high-quality person working for or under him for very long, but Snapchat is valued at $12 billion. Thus, the startup market is prone to the same talent inversions (i.e. low-quality people forming politically strong teams, and vice versa) as the corporate world. Hedge funds might be more predictable, in that many people have track records before starting them, but I also think that such people generally aren’t looking to hire external top talent (like me) opportunistically; they want people who’ve already got proven strategies. So that game is never going to be more than a niche. The need for “a track record” to break into it means that its growth is destined to be slow.
How about open allocation? I think that open allocation, properly deployed, solves the winning teams problem. Done right, it puts the employees all on the same team. Instead of jockeying for positions on the best teams, they’re all trying to make that one team (the company) succeed. If you could take 20 or 50 or 200 creative minds and create a “no justification zone”, meaning that people allocate their time as they see fit and don’t have to justify their own work– obviously, you want to encourage them to share it when it’s finished, but that’s different from the humiliating arrangement of having to justify one’s own working time in terms of, say, “sprints”– you could create a company where people don’t need to “get on” winning teams in order to have careers worth speaking of. I doubt that open allocation provides perfect meritocracy, but it’s a giant step in the right direction. Not only that, but I’d bet that open allocation teams don’t have the same sub-linear scaling (in which a team of 32 only 6-8 times as effective as an individual, and 250 only gets 16-25x) that we see among programmers.
At any rate, this is an important thing for me to think about. I’m getting to the age where it’s not really socially acceptable for me to just “be on” teams; I’m expected to build and connect them. Obviously, if I’m going to do that sort of thing, I want to do it right. I wouldn’t want to build a team and then play it to a loss. While it is possible to achieve much without paying much attention political success, it’s hard to keep achieving in the corporate world without rock-solid political game, because one’s time, people and resources are at risk of being paperclip-maximized into some other, lower-value effort. Now, I’m better than most of the people out there at spotting talent, which means that the first and hardest part of that process, I can perform. Could I drive a team to political success, when much of what I’ve seen is disaster? That’s more of an open question, to be honest about it.
