Comment by gaelian | Hacker News Reader

Comment by gaelian | Hacker News Reader

Parent Full thread

I met Mike Cannon-Brookes at an event that I attended some time ago and I happened to have a reasonably lengthy and interesting chat to him about - amongst other things - the culture at Atlassian. It happened to be the same week when GitHub had that PR disaster related to the apparently inappropriate treatment of one of their staff. I discussed this issue with Mike and asked if Atlassian had a similar view on the efficacy of the flat company structure that Valve, GitHub, et al espouse. He said that no, this was not the case at Atlassian because he felt that this kind of flat structure contributed to a poorer culture in the long run.

Obviously no one and no company is perfect and this story is a case in point, but I came away from that conversation with a much higher opinion of Atlassian than I had previously. It certainly seemed to me that Mike was serious about trying to cultivate a good culture at Atlassian and he also seemed like a genuine, no bullshit kind of guy.

I've also happened to have dealings of the professional networking/information exchange kind with one of the staff at Atlassian and was again impressed by how helpful that person was and the way they conducted themselves.

So in my anecdotal but first person experience, I would say that the culture at Atlassian is probably above par when compared to other similarly sized companies.

I discussed this issue with Mike and asked if Atlassian had a similar view on the efficacy of the flat company structure that Valve, GitHub, et al espouse. He said that no, this was not the case at Atlassian because he felt that this kind of flat structure contributed to a poorer culture in the long run.

Events like the Github fiasco should not be mentioned in the open allocation debate, because it's not like they're rare in a closed-allocation company. They're just as common, if not moreso. It just doesn't surprise us, when that shit goes down at a closed-allocation company, because open-allocation companies are held to a higher standard-- because open allocation companies are better.

Open allocation isn't a panacea. The Github event established that, if it wasn't already obvious. Open allocation isn't perfect. It is, however, superior to closed allocation is almost every way (at least, in software). Closed allocation only works better under severe time constraints.

"Flat organization" is somewhat different and usually (like "unlimited vacation") undesirable. Just as "unlimited vacation" means the policy is undefined, "flat organization" usually means the hierarchy is undefined-- which is dangerous, because "unofficial managers" are more likely to compete with those below them than official, titled ones.

In reality and perhaps paradoxically, true open allocation requires that people with credibility and leadership skill keep the company open and fair, which means that it requires (lightweight) management. It's not the Traditional Dickhead Management (i.e. you subordinate to this assigned person, whose goals may conflict with the company's and yours, or you are fucked) that you see in 90+ percent of companies, but it's not "flat" either. "Flat" is like "anarchy"; not stable. Open allocation, however, is the absolute right way to go.

I have read your post several times. I still have no idea what you're trying to say.

Valve and Github are well-known for open allocation, which is different from "flat organization". You can both, either without the other, or neither.

"Flat" organization is unstable and therefore probably not desirable. Open allocation (people can transfer freely as they wish) is a good thing and the only way to function at high levels as a technology company.

> Events like the Github fiasco should not be mentioned in the open allocation debate ...

Well I guess you'd have to take that up with Mike as you're really referencing his opinion not mine. In that instance I was neither trying to praise or criticise the kind of company structure that appears to be represented by Valve/GitHub (if "open allocation" better describes this phenomenon then I'll take your word for that). In fact I only mentioned it because it was the context within which my discussion of Atlassian's culture began on the evening in question.

I personally find this open allocation concept quite fascinating and think it potentially holds a lot promise. Most people working in corporate environments today would be lying if they said there wasn't huge room for innovation and efficiency gains at their workplace. To my knowledge however, I am unaware of any situation where an open allocation culture has been successfully retro-fitted to an existing corporate environment (if you know of one, I'd be interested). I'm also skeptical in relation to how well an open allocation culture would work out once a company grows beyond a certain inflection point. I'm not well read enough on this matter to put a stake in the ground and define that exact inflection point. But for the sake of illustration, let's say I'd be interested to know how such a culture would work out when applied to an organization consisting of thousands of employees as opposed to dozens or even hundreds.

To my knowledge however, I am unaware of any situation where an open allocation culture has been successfully retro-fitted to an existing corporate environment (if you know of one, I'd be interested).

That is hard as hell, to be honest. Paradoxically, I think it requires a top-down commitment. Companies usually want to try this sort of thing incrementally. So they put their best people into an R&D group that has a fair amount of autonomy. However, over time what seems to happen is that it's resented and attacked, and eventually has to spend more time fighting political battles than innovating. At that point, it's often just as underperforming as the rest of the organization, the experiment is judged a failure, and that autonomy is taken away. It's the "fish frying" problem (http://www.strikemag.org/bullshit-jobs/).

I'm also skeptical in relation to how well an open allocation culture would work out once a company grows beyond a certain inflection point. I'm not well read enough on this matter to put a stake in the ground and define that exact inflection point.

Gore-tex uses open allocation and is large (almost 10,000 employees, if I recall correctly). What I don't think OA can survive is too-rapid growth. I think it can scale to thousands of employees. I don't think it can tolerate personnel growth at the 100-150%/year levels that VCs want to see.

> Gore-tex uses open allocation and is large (almost 10,000 employees, if I recall correctly).

Thanks for the tip on Gore-tex, hadn't heard about them.

> What I don't think OA can survive is too-rapid growth. I think it can scale to thousands of employees. I don't think it can tolerate personnel growth at the 100-150%/year levels that VCs want to see.

So the important thing is to ensure that there remains a solid core of staff of an influential enough size that understands the open allocation process/culture in order to successfully orient new comers arriving at a reasonable pace. That's potentially quite an interesting insight, I'll remember that one.

Reply on news.ycombinator.com