Why does a company full of smart people make stupid decisions? How do we keep it from happening?
The natural tendency may be just to make decisions that are acceptable and least-common-denominator; decisions that no one disagrees with -- because it's easier and less contentious. But this doesn't seem to be the path to success.
Microsoft is an easy target right now, but it's because they're such a puzzle. Filled with really smart people, effectively infinite financial resources. And buffaloing themselves right over a cliff. I finally figured out that the reason Bill Gates used to loudly proclaim that innovation is what Microsoft is all about is not because he actually believed it, but because he wanted it to be that way. He saw that the only way the company could survive is if they actually became innovative. And he got out because he eventually understood there was no way to turn the ship away from the iceberg.
The company that avoids collective stupidity is a statistical anomaly. Something about the way we do and think about decisions -- something so fundamental that we do it below the threshold of consciousness -- builds in this inevitable tendency towards the non-innovative, non-creative, non-fun organization.
I think one issue is desperate hiring practices which come from forced growth. You "must" grow, so much so that you start saying "we need a body to fill this position," and so the "people are our most important asset" maxim goes quickly into the trash. You start hiring people that don't have the company vision, that don't fit into the team, but they have "qualifications" and they are available and you've got this project that needs to be staffed up right now because there's a contract and money, etc., etc..
And those people eventually start disagreeing with decisions that would once have been clearly seen to be in the company interest, and to placate them those decisions are dumbed down and pretty soon you've got collective stupidity.
Another aspect of this is the size of the group. One of Hewlett-Packard's fundamental tenets was the idea that no unit of the company should be larger than a small town, so you could know everybody. If a unit got too big, they split it to keep things small. It worked pretty well. It just seems like when a group gets large enough, you have the mediocritization effect kicking in even if everybody in general seems to be pulling in the same direction.
So perhaps the answer is to keep things small; keep the units of your business broken up into teams where everyone gets along and moves forward together, and evaluate the health of the business based on the health of those individual teams. Maybe we are hard-wired to be unable to operate optimally at anything larger than tribe level. Eric Raymond once told me he was going to write about what he considered the future of the software business, which would be independent teams that come together to work on a particular project, and when the project is over go back to being individual teams again, joining up with different teams for the next project.
These are just my thoughts on the problem. What do you think causes collective stupidity, and how can we design an organizational structure to prevent it?
The larger an organization becomes, the less connected an individual is to the external consequences of an action such as won or lost business, increased or decreased support costs, lawsuits, etc; and the more connected an individual is to the opinions of others.
This creates strong incentives to follow the herd, because there is no personal risk in following the herd, at least until the company goes belly up.
I'd also say that the major way that companies get off-track, which is by switching their focus from customers to stockholders.
A company's REAL constituents are customers, but when they stop trying to please them, and instead try to please stockholders, or worse, "analysts" who set unrealistic expectations based on other unrealistic expectation, then poke them with a fork--they're done.
This is what has happened to every major tech company that's either gotten off-track, or completely run aground. Microsoft simply has so much momentum that they can be off-track for years before they grind to a halt, and unless they get back on-track, even they will.
WHO IS THE CUSTOMER. WHAT DO THEY *REALLY* WANT AND NEED. "Really" is the key word. Not what you think they want. Not what you think they should have. But wehat they really want and need (and not what some small, vocal group says they want).
There's also the very basic mistake of trying to get people who hate your product to love it, and by changing it to try to make them love it (which they never will) you end up alienating the people who did love it in the past.
I've been a long time fan, in fact - I read your books while going to school. Just wanted to say thank you for all the work you have put into the software engineering community.
I've been working now in software for nearly 8 years, steadily moving into larger and larger companies. I often feel as you do, (a fish lost in a ocean of people), and have seen my share of corporate waste. I have also seen pockets of innovation which struggle to get the visibility they need.
I think in general, people whom are software engineers need to step up and lead more. Explain to management and other engineers WHY we are building something, WHAT problem it is we are trying to solve. For too long, the Java community (and others) have been too focused on "Thats a cool framework".
I believe leading & establishing innovation communities within corporations is the only way to really solve the problem. To pull this off, companies have to recognize the need, and get the right people in the right places. If we don't act soon, eventually companies will be out of resources.
If I understand your take on stupid decisions right, it is decisions likely to damage a company's business in the long run. But these are not necessarily stupid decisions. It is what owners of companies want: maximizing profit in the short run. Stock owners have a different point of view on when a company is on track than technology people. If a company makes a huge profit this year, allowing them to sell their stock at insanely high prices, it doesn't matter that the company will die next year because all of its customers will be pissed off. IMO, this is the primary reason for what you call collective stupidity. Do you think that engineers at MS are making decisions? Their decisions are driven by sales and marketing, and for sales and marketing lack of standards compliance, incompatibility with other systems and proprietary technology are strategic commandments not set by engineers. Which is exactly what makes me, a programmer, think that MS is off track. Microsoft, the easy target, is holding on to its strategy of grabbing their customers by the balls and squeezing them for money. As long as it's feasible, it's a smart business decision.
Microsoft set aside, and taking only really stupid decisions into account, I think Peter's principle has more to do with it: eventually, in a large enough organization, more and more decision making positions will be filled by people underqualified to make decisions required by the position they occupy. Which will promote even more stupid people upwards. Eventually, management looses touch with reality, and no matter how frustrated the people actually doing the work are, bad decisions are made. From this point of view, size surely does matter. Nevertheless, I don't think it's a matter of all or most people in an organization catering to the least common denominator, or to decisions not in the spirit of the company's culture. It's just the upwards promotion of people not being able to make the right decisions.
Yet another point of view: HP, Siemens, Sun Microsystems, Adobe were all founded by technology people, not by business people. They were quite successful and stayed on track for a long time - from a technologist's point of view. Then they started to get off track. IMO, this happened the moment when technology people were removed from decision making, and most decisions started to be taken by economists, sales people or marketers. Take Apple: Jobs was pushed out by non-techies, and the company immediately started to fall appart. The return of Jobs managed to turn the company around. The same with Microsoft: Ballmer was the first business manager hired by Gates, and managed to get the company off track - from a technologist's point of view. From Ballmer's point of view, Microsoft is doing just fine. In spite of lots of negative signals, Microsoft still earns lots of money, has a larger installed base than any of its competitors (OS, Office suite, browser, media player), advances slowly but steadily in some areas - like for example eating into Notes/Domino mail and collaboration market, and getting serious companies build their application server infrastructure on .Net instead of Java. Why would Ballmer be unhappy? How could you qualify his decisions as being wrong, when judging them by his criteria, which is just making money? Ballmer himself would probaly turn the ship around, and would perfectly be able to do so, if he saw the iceberg comming. But from his point of view, there's no iceberg close enough to start to really turn the ship. Of course, he could cater to pressure from engineers and technology people outside, but would this increase his profit? He probably just calculated that the stock price decrease of Microsoft caused by various business practices perceived as being not OK by the public is less of a loss than the loss that would result from a change in strategy, making MS products and platforms more open and standards compliant.
Making an analogy to the biological world: dinosaurs will eventually die out. It's just a matter of evolution. Nevertheless, by their nature, they have a long life span.
My feeling based on my experiences is that a lot of this has to do with management succumbing to the flattery of their subordinates. Instead of promoting those that question their decisions, they promote those that always agree with and 'support' their questionable decisions. This creates an atmosphere of cult-like belief in absurd ideas.
I see high-level staff repeating completely ludicrous ideas and assumptions and making decisions based on them all while everyone in the trenches knows that it's nonsense. Fragile egos and the kind of arrogance only the ignorant possess. That's my theory of the source of these problems. Another problem I've seen is a culture of hammering anyone who makes a mistake. This leads to people putting a lot more effort into hiding mistakes and throwing good money after bad than into doing makes sense.
I think it's because companies are insular in nature, and the longer one is in the company, the more skewed their view of the real world is. Companies tend to stabilize things, (to have repeatable process, etc) so over time become a less challenging environment than the real world for the people who have been there a long time and have learned the politics of the company. These people tend to be managers and lifers (hoping company won't fail before they retire).
In this safer environment, the instinct is to hold steady. And that means not introducing new technology as rapidly as in the real world (which means you fall behind eventually) and not bringing new ideas as rapidly as in the real world (which means you fall behind eventually), and ultimately you have old technology and old thinking and that will not produce solutions to real world problems.
This is why startups are more brutal but more innovative. They live in the "current real world month" and have to deliver valuable solutions to actual current problems.
So, collective stupidity happens when people try to apply organizationally-reinforced old thinking and old tech to current problems.
One reason for this is, that people tend to make decision by induction: baseing conclusion upon past events. It's a natural thing to do for humans: I expect my pen to go downwards if I let it go, because it always has done so. Often , it's a very effective way of reasoning. But not always. For business people it is a very dangerous way of decision-making. Business is not gravity. "We'll do it like we always did it, because it has always worked and we made millions this way". yeah right... "älways" being 20 or 30 years, not 70.000 or whatever years for man to exist on this planet. Software is in its infancy, guys ! Remember Bertrand Russell's turkey: it expected the farmer to come by with food each morning, living a happy, fattening life for a year, not knowing about Christmas... I've met ICT business people reason by induction. Bugging their code to unmaintainability, offering many employees and their families a livelyhood, based on things gone by.
> > Part of the problem with common inductive reasoning is > > that it is done based on singular occurrences, rather > than > > repeated ones. > > Another problem is that in many cases it works. The trick > is to be able to see when it might not.
What you said about Bill Gates is very interesting and really resonates with me. I sometimes feel that what you describe as collective stupidity starts at the top of an organisation and filters down the ranks.
So, if the CEO has a particular strong idea about something, his direct reports feel that they can't really disagree, and so they 'tow the line' so to speak. Of course their direct reports then end up feeling the same way, and this slowly filters down, to the lower level of an organisation. That people at that level don't really need to tow the company line, but they don't feel like they have the power to drive change.
This is more of an organisational culture problem, and is possibly one of the symptoms of forced growth.
If you can foster a feeling within the company that, people can drive change and that 'towing the line' is not only unnecessary but discouraged, I think it won't matter how big the units within the company are. Doing this of course is the real challenge, and hiring practices must play a major role. I don't know of anything else that can have as big a detrimental effect as having the wrong people in key roles. Of course the bigger the company is the harder this is to avoid, hence collective stupidity becomes more likely.
This topic is of particular interest to me as I work for a company that is experiencing growth and is expanding our development organization. We are trying to do it in ways that scale in the sense that we do not want bottlenecks in production or innovation.
Another danger of growth we are trying to avoid is the loss of accountability. Bruce and many of the follow-ups seem to be stabbing around the heart of this.
I believe maintaining accountability as low in the organization as possible - in the hands of the technologists as one post put it - is a key to avoiding organizational Alzheimer's.
There are many other factors that dovetail with this. For example, we have to hire the right people. We specifically look for technically strong people with integrity, initiative and the ability to communicate and collaborate well. We seek out those who are comfortable with change, managed risk-taking, and have technical leadership potential.
These are the people who eventually lead our small teams of developers and who are accountable for delivery of quality product. We managers are responsible for enabling these leads, coordinating their projects, fostering communication, and promoting individual and team growth.
It has been a struggle to stay on the hiring "moral high ground" in the face of executive pressure to grow. At some point, our positive results gave us the break we needed. The executives saw the value in our strategy and, although they really, really want us to grow faster, they allow us to grow in a good way. This is the power of trust as a tool for greater freedom.
There are many ways to fight off the plunging organizational IQ. To me, accountability at the lowest possible level is a key component.
I think it is because of mediocrity. If people know they are the best and valuable by the company, irreplaceable or can easily find another job, they will say "i think this is BS" if they see it. If they become dependent on their position, or lazy about changes, they will remain silent.
Flat View: This topic has 16 replies
on 2 pages