Joel Spolsky's business operating system | Hacker News Reader

Joel Spolsky's business operating system

pixelmonkey.org

93 pointspixelmonkey17 comments

I agree with the central point: there's much evidence that Joel Spolsky knows what he's doing, and executives would be wise to implement his suggestions.

I've talked to engineering executives who agree, "I've read all of Joel's stuff", but then fail to implement his ideas at a concrete or abstract level. I don't know why.

A lot of managers, and even many executives, are not in the position to implement Spolsky's ideas. Take his suggestions on office layout, for example. If you occupy the C-suite at any given company, or if you're the founder of your own company, you might be able to call shots on things like that. If you're not, you probably can't. You probably have a boss, who has a boss, and you have bureaucracy to wade through.

Joel's advice is best taken at the founding stages of a company. It's extensive, and it's structural, and it needs to be part of the company's foundation. If you're the one group or division at a large company that "goes rogue" and operates on Joel's principles, even if you have the political clout to do so, you're painting a target on your back. You're taking a huge political risk. Maybe it pays off. Maybe it doesn't. At any rate, very few people reach executive status at established companies without a keen understanding of internal politics, and the constraints it imposes.

I'm not defending politics, of course. I'm just saying politics exists, and it's a powerful organizational force.

If you successfully and reliably deliver a higher quality product in a shorter time-frame than "the old way" one of two things will happen.

1) A whole lot more work and responsibility will come your way.

2) Other divisions will want to emulate your success.

If you are not successful, then yes I think sticking your neck out in a big organization is a good way to lose your head. The thing to remember is that if this organizational change is well executed then your likelihood of success increases.

I'd argue that there is also a 3),

3) Other divisions will get envious and seeing no way to emulate you they will try and stab you in the back or in other ways sabotage your work.

Maybe they will fail, if you're that good and/or have great political skills. But it's definitely something that might happen, and you have to be prepared for, and potentially prepared to lose.

Yes, but of course, it's never as simple as pressing a button and implementing the ideas. You have to get buy-in for the ideas, presumably one idea at a time. You have to justify, in tangible and financial terms, what the cost/benefit of making the change will be -- especially because implementing any structural change will require a very large, very real cost up front. And the problem with one-idea-at-a-time is that each idea, in isolation, carries a high political risk, but less potential upside. And so the benefit of implementing a Spolsky-like program really accrues when implementing quite a few of the concepts at once. Very few people have the juice to do that. Furthermore, something could easily go wrong, entirely unrelated to one's Spolskification program, but which torpedoes the program nevertheless.

Another powerful force to consider is employee morale. Employees tend to settle into a routine and resist change to that routine -- even if it's nominally better for them. You'd be surprised at how powerful a force this is. I can almost guarantee you that if you went to any big company with an open-layout, cube-farm floor plan and proposed that everyone be given their own office, a significant percentage of those employees would grumble about it. It sounds silly to think so, but it will happen. People fear change. They tend to read into it. They think it's a harbinger for more sinister or scary things to come. Loss-aversion is very real and very strong.

Of course, the upside to making big changes is the chance at big success. If you turn around your division, or if you're the one division outperforming all the others, you're a hero. And other divisions will follow suit. That's a wonderful thing. But the risk that that won't happen, and that your changes will backfire or be scuttled in some way, is every bit as real as the potential upside. When faced with that calculation, and especially when taking into account the stability of a very nicely paying and high-ranking job, a lot of managers and execs will opt not to rock the boat.

Again, this is not a defense of change-aversion. It is an explanation for why you don't see more change in established firms. The vectors of big change tend to come from outside: a new CEO, a new executive hire, or the formation of a new business or division. In those circumstances, the organizational and psychological frame is oriented towards change, rather than towards the preservation of the status quo. The person leading the change has a clear mandate to do so. In other, more day-to-day circumstances, the mandate is less clear and less easily secured.

Need to develop a viral version of Joel's OS.

Spolsky is a great writer and has attractive ideas, but there are so many counter-examples: companies who violate one or more of his rules/preferences, yet have achieved greater success (measured however you want: financial, cultural, technological, etc). It makes it hard for to me to see his advice as necessarily better or smarter than anyone else who is running a company.

I've been reading Spolsky's writing since he began, and it's strange I still think of his company as having 10 or so employees.

Honest question: what do the 30+ people on the production side of Stack Exchange do? This frankly seems like a huge team to me. On the other hand, maybe that's what's required for sane scheduling and clean code. My experience is with 1-5 devs doing any mad hack or taking known-shortsighted actions to meet the current deadline.

If this article from HighScalability is any indication, it looks like the 30+ on the product side of Stack Exchange are running a very tight ship:

http://highscalability.com/blog/2014/7/21/stackoverflow-upda...

"beautiful daylit private offices"

Although private offices may be more difficult to achieve, I'm amazed at the number of companies that do not take daylight into consideration. It has been proven to increase worker performance and mood, yet companies continuously rent office space in dark, cavernous buildings. Sometimes they even brag about it. Then they wonder why they can't hire decent workers or why their workers are leaving or simply being unproductive. No, the small glare on monitors is not a reason to rent some dark warehouse with no windows; it's an excuse, a pretty lame one at that. Of course, this is one of the many problems solved by working from home.

This is really cool. I think I'm going to start blogging about our business practices in a couple years, which are very different than most companies. In some sense, we are running our company more like a distributed project, like Linux or Bitcoin.

Our company's mission is: "Empowering individuals, uniting communities."

Our company's motto is: "People live lives, companies create products."

Why not something like "don't be evil?" Because our company's distributed, and focused on creating products that empower people. The whole thing about a centralized company opposing its users or doing something to screw everybody is exacerbated when there is massive centralization, data centers, a building with lots of employees, jobs entrenched interests, lobbying, etc.

We are embracing the new distributed world that git, bitcoin, etc. help usher in. So company's mission involves build products to empower people. Will the company be worth billions of dollars? Probably, because we're going to be the first movers on our platform. But the platform powering what we do will create tens of billions of value outside what we capture.

I think this new movement to decentralize our institutions really does help empower individuals and unite communities. And it's gaining steam along the lines of http://asyncmanifesto.org/#.U44s7gov84U.gmail and https://www.ted.com/talks/clay_shirky_on_institutions_versus...

I wrote this about 3 years ago: http://myownstream.com/blog#2011-05-21

"I think I'm going to start blogging about our business practices in a couple years"

You should do it now! Why wait a couple of years? Even if you don't know everything now (who does?), I'm sure I speak not only for myself when I say I'm always interested in new ideas how to structure or run companies.

I already do, but not regularly: http://qbix.com/blog

> Will the company be worth billions of dollars? Probably, because we're going to be the first movers on our platform.

You crack me up, Greg.

"Don't be evil" was memorable and conveyed an ethos. It was an explicit poke at Microsoft, the evil empire to hackers of its time. By comparison, what you wrote is extremely generic and fails to pack a punch. It's hard to imagine anything more flaccid than a motto based on a process that has increasingly become the convenient default of tech companies. Joel's motto isn't about the process (distributed, asynchronous, etc) but about the people, and it's anything but convenient to implement.

Don't worry Per. I am not that pompous as you would like me to be :)

Many companies are trying not to be evil either.

We don't have to worry as much about doing evil things.

PS: How's life?

Sorry about the timeouts on that page for the first 20 minutes -- turns out my Wordpress page caching plugin had gotten disabled somehow. Fixed now!

I would be happy if I worked for a company with just the high end coffee, let alone everything else.

Reply on news.ycombinator.com