This is the second of a weekly series of posts on various aspects of social software design I find interesting, here is the full list. Each of these posts are written over the course of a few hours in a straight shot. Contents may be mildly idiosyncratic. To vote on what I should write about next, go to this Quora question.

Ice cream sundae

The people who most want to meet people are the people who the least number of people want to meet. The people who are the most desperate to date are those who the least number of people want to date. The people who are the most eager to talk are the ones who the least number of people are interested in hearing. It is the ignorance of this fundamental principle that I see at the heart of so many failed social software designs. This is what I call the Evaporative Cooling problem and one I believe must absolutely be tackled head on by the designers of any communal gathering product unless they want to see their product descend into a squalid lump of mediocrity.

The Evaporative Cooling Effect is a term I learned from an excellent essay by Eliezer Yudowsky that describes a particular phenomena of group dynamics. It occurs when the most high value contributors to a community realize that the community is no longer serving their needs any more and so therefore, leave. When that happens, it drops the general quality of the community down such that the next most high value contributors now find the community underwhelming. Each layer of disappearances slowly reduces the average quality of the group until such a point that you reach the people who are so unskilled-and-unaware of it that they’re unable to tell that they’re part of a mediocre group.

Evaporative Cooling is a dynamic that can apply to both real world and online communities but the affordances of the Internet make it particularly susceptible to Evaporative Cooling. By looking at real world social structures, we can get some clues as to both what causes Evaporative Cooling and what are effective ways of preventing it.

Example the first:

Moving to San Francisco, it was amusing to me, unearthing the social structures around networking that go on here. There is the public “scene” of parties, events & mixers. Alongside this is an entire shadow community of private, invite only, exclusive events which is where all the real work in the Valley is done. It is possible to live your entire life in the Valley, wandering around amicably being blithely unaware of the shadow ecosystem. You could go to the same events every week with the same mix of aspiring entrepreneurs, social media marketers, CEOs of dipshit companies, bloggers & the occasional A-Lister who is forced to be there out of professional obligation.

But, if you’re halfway decent and capable of networking, you’ll soon find yourself with an entrée into a small part of the shadow economy. How far down the rabbit hole you choose to go is purely a function of your innate function and drive. For every layer of exclusivity, there’s almost certainly one more exclusive that you’re not aware of. Some of these venues are well known; TED, Davos, Sun Valley. But for every one of these you’ve heard of, there’s certainly at least a thousand more equally as exclusive gatherings you haven’t. After a while, you start to subscribe to what I call the Groucho Marx rule. You stop attending any event which would have you as a participant.

Lesson the first:

Openness is a major driver of Evaporative Cooling. If anyone can join your community, then the people most likely to join are those who are below the average quality of your community because they have the most to gain. Once they’re in, unless contained, they end up harming the health of the community over the long term. Communities that are allowed to select their members in some way are much more immune to Evaporative Cooling. Unfortunately, most viable internet businesses have no choice but to set their business model to open. The nature of most Web 2.0 businesses is that they depend on extracting a tiny bit of value from a large number of users and are betting on their fuck you exit from massively exploding in scale. Building a thriving community that tops out at 10,000 members over the course of 10 years isn’t going to pay the bills.

Example the second:

One of the communities that I’m part of down here is BayCHI. It’s a community that’s been around for 20 some years now and the quality of the talks and people who attend is still excellent. It seems to have only minimally succumbed to Evaporative Cooling. Why is this? A large part is due to what I call Social Gating. Social Gatings are mechanisms that allow participants to self-select out of the group. In the case of BayCHI, the social gate was the nicheness and unglamorousness of the content. The only people who would choose to participate in this group in the first place are those who find the talk sufficiently interesting to take 3 hours out of their life. This, by itself set a minimum bar.

Lesson the second:

Social Gating is a powerful force and, unlike direct exclusion, works in a much more scalable fashion at Internet sized growth rates. However, it is also a much more subtle one and requires a deft hand to get right. Nicheness is just one possible social gate, charging money is another popular one. But there are an entire constellation of more nuanced ones. Spelling, for example, is an interesting social gate. Just seeing a forum in which ppl spel liek thiz instantly polarizes you onto one side or the other. At the other extreme Quora, in it’s very early days had an incredibly Orwellian system in which Quora staff would routinely directly edit the contents of your answer to fix spelling and grammatical errors. I’m planning to dedicate an entirely separate Social Software Sunday blog post to Social Gating so stay tuned (pro tip: If you want to see it faster, go to Quora and add it to the list).

Example the third:

Another event that I attended this week that had a remarkably high quality of participants was Warm Gun. Among the people in the room were the Director of Design at Facebook and the Director of Design at Google. How did Dave McClure get these two in a room? He put them on a pedestal, literally. They were invited to take part in a panel discussion on how designers & engineers could better work together and it was the inducement of special treatment that made these very busy & high value contributors deign to be in the same room as us design peasants.

Lesson the third:

Unequal roles of participation can help shift the gradient of power and kill the evaporative cooling. When the community is small, such processes can be managed through the social layer. High value participants are treated as special because they have recognition & reputation from the community. But, as the community scales, these social mechanisms break down and often, if nothing is done to replace them, high value members get especially miffed at the loss of special recognition and this accelerates the Evaporative Cooling.

Explicit reputation systems like karma are probably the most popular way online communities have implemented unequal roles. But, for some reason, online communities seem particularly resistant to the type of elitist promotion structure common in real world institutions. In Academia, high school students have to fight to become undergraduates. Undergraduates have to fight to become PhD candidates. PhD candidates have to fight to become adjuncts. Adjuncts have to fight to become tenured and tenured professors have to fight to become Dean. I can’t even think of a single online community that bears even the slightest resemblance to this sort of power structure. This is something to ponder for a later piece.

Example the fourth

Finally, I will examine what I consider to be one of the most successful technological systems ever at scaling while maintaining quality: Facebook. I joined Facebook when it was less than a million members. Since then, it’s managed to grow by a factor of 500 but the quality of my experience has dropped by only maybe 50%. The reason why is because when some random person is participating in Facebook from Brazil, it has an absolutely negligible effect on my experience. Because every user only ever see their tiny corner of Facebook, every user is in direct control of their own experience. Lest you think this is a property that is intrinsic to Social Networks, Orkut was brought down precisely by those random people in Brazil. Facebook’s design, especially in the very early days, was especially conscious of this design dilemma and designed around it masterfully.

Lesson the fourth:

There are two fundamental patterns of social organization which I term “plaza” and “warrens”. In the plaza design, there is a central plaza which is one contiguous space and every person’s interaction is seen by every other person. In the warren design, the space is broken up into a series of smaller warrens and you can only see the warren you are currently in. There is the possibility of moving into adjacent warrens but it’s difficult to explore far outside of your zone. Plazas grow by becoming larger, warrens grow by adding more warrens.

These are the two fundamental patterns of social spaces. Every social space can be decomposed down to a collection of plazas and warrens. In Facebook, your profile, friends and newsfeeds are warrens but fan pages, groups & events are plazas. Twitter is mostly a warren with the exception of trending topics which is the one plaza. On forums, the front page and topic listings are plazas but each forum thread is a warren.

Plazas and warrens both have their unique set of tradeoffs. Warrens are notoriously difficult to get started. New users, stuck in empty warrens often don’t know how to connect to hubs of activity. The onboarding process is crucial and still not well understood (Friendfeed found that people needed to add at least 5 friends to have a reasonable chance of sticking with the service). On the other hand, plazas only need to be started once and then they remain a hive of activity for new users to participate in from the first day.

Plazas are much more visible than warrens so it’s easier to watch and understand your community. In communities, like in justice, sunlight is often the best disinfectant and the neglected spaces often become thriving breeding grounds for all sorts of social pathologies.

But the one absolute killer feature of warrens is that they allow your community to become almost perfectly scale free and grow like mad without ever sacrificing quality. This alone, makes them a design element that’s heavily worth studying to figure out what are the good social designs.

It’s also interesting to note that the real world is intrinsically warren while the online world is intrinsically plaza. In real life interactions, the physics of sound mean that we can only ever talk to a few people at once. Every person gets a “personalized” social life. To give every person the exact same content takes special work. Online, the easiest model to program is to serve the exact same bits to every requester. To provide “personalized” content takes special work. It is interesting to observe how this difference has influenced the evolution of these two mediums.


Evaporative Cooling is a fundamental social dynamic and one that is corrosive to the long term health of communities. This post contains barely 1% of everything I could write about Evaporative Cooling but I’m already at 2000 words and I’m not looking to write a novel here. They say ideas are worthless and execution is everything. Since I’ve gotten to the Valley, I’ve heard probably close to 100 pitches for social products in random conversation. About half of them involved a meeting place dynamic of one kind or another and about 80% of those, as they were conceived, would be killed dead by Evaporative Cooling. It is absolutely essential if you’re to be designing a social product that you deal with this issue up front or you’re just a dead man walking.

To be notified of the next Social Software Sunday piece as it’s posted, you can subscribe to the RSS feed, follow me on twitter or subscribe via email:





  1. pyridium dosage says:

    November 2nd, 2011 at 7:07 am (#)

    This is the second of a weekly series of posts on various aspects of social software design I find interesting,

  2. website registration india says:

    November 2nd, 2011 at 7:34 am (#)

    Yes, i found your post most interesting one as well as useful.. 
    I will be waiting for other interesting posts from you in the nearest future.keep it up.

  3. Joomla Development says:

    November 2nd, 2011 at 9:52 am (#)

    Wow, really very tasty it is! I feel like to have it! Thanks for sharing!

  4. designerhandbagsoutlet says:

    November 8th, 2011 at 11:32 am (#)

    Thanks for your opinion. discount handbags I totally like with it.I like cheap designer handbags as well and someone is looking for cheap handbags People usually prefer fake designer handbags, especially when they visit some handbags outlet. In some areas,people like to wholesale clothing. They will find some store doing very cheap or discounted Chanel Handbags outlet. Certainly,  is a good way to earn money.
    designer handbags outlet

  5. diploma financial planning says:

    November 22nd, 2011 at 6:43 am (#)

    thanks for the comment

  6. medion akku says:

    November 24th, 2011 at 8:00 am (#)

    Phenomena is plural – phenomenon is the singular.

  7. MSI Megabook S271 akku says:

    November 25th, 2011 at 7:40 am (#)

    Phenomena is plural – phenomenon is the singular.

  8. domain register says:

    November 29th, 2011 at 11:31 am (#)

    Great!  Well, I admire the precious time and effort you put into it, especially into interesting articles you share here! It was very interesting..  loved it

  9. Bangalore website design says:

    February 16th, 2012 at 4:28 am (#)

    I really enjoyed this site. This is such a Great resource that you are providing and you give it away for free. It gives in depth information. Thanks for this valuable information.
     Bangalore website design  

  10. web designing company says:

    February 24th, 2012 at 11:32 am (#)

    Really a cool information is presented ..thanks for this blog here.

  11. Building Castles in a Sociological Sandbox | OnTheSpiral says:

    May 19th, 2012 at 12:12 am (#)

    […] meeting spaces.  We commingle in these digital environments for only a period of time before evaporative cooling pushes us […]

  12. Evaporative Cooling – or “overrun by muppets” « says:

    May 19th, 2012 at 11:07 am (#)

    […] that this is what I thought about at all when I read this highly entertaining and thought-provoking post about (online) social networks. The Evaporative Cooling Effect is a term I learned from an excellent essay by Eliezer Yudowsky […]

  13. Quora says:

    May 29th, 2012 at 7:45 am (#)

    What are some good images of Spring in the home country of Quora users?…

    If I’m the only one on Quora who is posting interesting/good answers, then that says more about Quora and it’s community, than it does about me. What she means, is that there is more to having a successful website, than just having loads of content o…

  14. Guest says:

    June 3rd, 2012 at 12:23 am (#)

    Orkut only SURVIVED because of those “random people in Brazil”. It’s adoption elsewhere was very low even before brazilians started flooding in.

  15. 社交网络中的“蒸发冷却效应” | 遠方 says:

    June 6th, 2012 at 11:40 pm (#)

    […] 硅谷一名社交网络设计师在他的博客中提出了这样一个有趣的现象:当新成员不断涌入一个开放社团时,社团中最有价值的成员会发现社团成员平均水平的降低让自己继续呆在这里已经没有意义了,于是他们就会选择离开。这批成员的离开进一步降低了社团的价值,于是恶性循环开始了:越来越多高价值的成员选择离开,直到有一天这个社团彻底的沦陷成了一个平庸的组织。 […]

  16. 社交网络中的“蒸发冷却效应” | | Mr.Sun Studio | Creativity PractitionerMr.Sun Studio | Creativity Practitioner says:

    June 7th, 2012 at 1:26 am (#)

    […] 硅谷一名社交网络设计师在他的博客中提出了这样一个有趣的现象:当新成员不断涌入一个开放社团时,社团中最有价值的成员会发现社团成员平均水平的降低让自己继续呆在这里已经没有意义了,于是他们就会选择离开。这批成员的离开进一步降低了社团的价值,于是恶性循环开始了:越来越多高价值的成员选择离开,直到有一天这个社团彻底的沦陷成了一个平庸的组织。 […]

  17. Y H says:

    June 7th, 2012 at 4:45 pm (#)

    Nice discussion. Similar things happen to all forums that aimed at a high level discussion. I think StackOverflow and MathOverflow are doing pretty well in the sense of maintaining a high standard.

  18. 社交网络中的“蒸发冷却效应” « 得看看-精品阅读|速查|分享 says:

    June 7th, 2012 at 9:32 am (#)

    […] 硅谷一名社交网络设计师在他的博客中提出了这样一个有趣的现象:当新成员不断涌入一个开放社团时,社团中最有价值的成员会发现社团成员平均水平的降低让自己继续呆在这里已经没有意义了,于是他们就会选择离开。这批成员的离开进一步降低了社团的价值,于是恶性循环开始了:越来越多高价值的成员选择离开,直到有一天这个社团彻底的沦陷成了一个平庸的组织。 […]

  19. 社交网络中的“蒸发冷却效应” | My Blog says:

    June 8th, 2012 at 5:27 am (#)

    […] 硅谷一名社交网络设计师在他的博客中提出了这样一个有趣的现象:当新成员不断涌入一个开放社团时,社团中最有价值的成员会发现社团成员平均水平的降低让自己继续呆在这里已经没有意义了,于是他们就会选择离开。这批成员的离开进一步降低了社团的价值,于是恶性循环开始了:越来越多高价值的成员选择离开,直到有一天这个社团彻底的沦陷成了一个平庸的组织。 […]

  20. 社交网络中的“蒸发冷却效应” | IT前线! says:

    June 8th, 2012 at 8:05 am (#)

    […] 硅谷一名社交网络设计师在他的博客中 提出了这样一个有趣的现象:当新成员不断涌入一个开放社团时,社团中最有价值的成员会发现社团成员平均水平的降低让自己继续呆在这里已经没有意义了,于是 他们就会选择离开。这批成员的离开进一步降低了社团的价值,于是恶性循环开始了:越来越多高价值的成员选择离开,直到有一天这个社团彻底的沦陷成了一个平 庸的组织。 […]

  21. 【SNS如何助力你成为牛人】谈一谈社交网络中的“蒸发冷却效应” | 【】找牛网 says:

    June 9th, 2012 at 1:24 am (#)

    […] 硅谷一名社交网络设计师在他的博客中提出了这样一个有趣的现象:当新成员不断涌入一个开放社团时,社团中最有价值的成员会发现社团成员平均水平的降低让自己继续呆在这里已经没有意义了,于是他们就会选择离开。这批成员的离开进一步降低了社团的价值,于是恶性循环开始了:越来越多高价值的成员选择离开,直到有一天这个社团彻底的沦陷成了一个平庸的组织。 […]

  22. 社交网络中的“蒸发冷却效应” | 曹志士 says:

    June 25th, 2012 at 11:22 am (#)

    […] 硅谷一名社交网络设计师在他的博客中提出了这样一个有趣的现象:当新成员不断涌入一个开放社团时,社团中最有价值的成员会发现社团成员平均水平的降低让自己继续呆在这里已经没有意义了,于是他们就会选择离开。这批成员的离开进一步降低了社团的价值,于是恶性循环开始了:越来越多高价值的成员选择离开,直到有一天这个社团彻底的沦陷成了一个平庸的组织。 […]

  23. Plazas vs Warrrens | Alvin Hsia says:

    July 10th, 2012 at 4:24 pm (#)

    […] blog post about the architecture of online social systems. As more niche and nuanced platforms emerge to take over specific use cases not fully satisfied by […]

  24. 社交网络中的“蒸发冷却效应” | IPOWall科技 says:

    August 19th, 2012 at 12:48 pm (#)

    […] 硅谷一名社交网络设计师在他的博客中提出了这样一个有趣的现象:当新成员不断涌入一个开放社团时,社团中最有价值的成员会发现社团成员平均水平的降低让自己继续呆在这里已经没有意义了,于是他们就会选择离开。这批成员的离开进一步降低了社团的价值,于是恶性循环开始了:越来越多高价值的成员选择离开,直到有一天这个社团彻底的沦陷成了一个平庸的组织。 […]

  25. Plazas vs Warrrens | Alvin HsiaAlvin Hsia says:

    November 7th, 2012 at 4:16 pm (#)

    […] blog post about the architecture of online social systems. As more niche and nuanced platforms emerge to take over specific use cases not fully satisfied by […]

  26. Four short links: 13 December 2012 - O'Reilly Radar says:

    December 13th, 2012 at 3:55 am (#)

    […] Evaporative Cooling — thoughtful piece about the tendency of event quality to trend down unless checked by invisible walls. (via Hacker News) […]

  27. On Social Networks: Plazas vs Warrens Alvin Hsia says:

    February 9th, 2013 at 10:27 pm (#)

    […] blog post about the architecture of online social systems. As more niche and nuanced platforms emerge to take over specific use cases not fully satisfied by […]

  28. says:

    February 12th, 2013 at 8:54 am (#)

    […] Social Software Sundays #2 – The Evaporative Cooling Effect « Bumblebee Labs Blog […]

  29. Quora says:

    February 17th, 2014 at 1:14 pm (#)

    Why do people stop using Quora?…

    I can think of three reasons. Quora is not a meritocracy, it’s an aristocracy On normal websites, the more content you post, the more positive feedback you get. If you post photos on Facebook about your life, you get likes and nice comments and ravid …

  30. Dadie Host says:

    May 14th, 2014 at 9:31 pm (#)

    I had the most aggravating experience in years with Quora. They didn’t
    speak good English, talked with very bad accents, talked too fast, did
    not comprehend what I needed, and worst of all, were stupid. I had had a
    very very good service person on with me three weeks ago, and he helped
    me perfectly. I had very high hopes when I called yesterday about one
    more simple issue, and got these idiots. I now wonder about Quora
    customer service ability. I ended up asking someone at work and fixed
    the problem myself after several minutes of research.

  31. The trough of social media disillusionment says:

    August 26th, 2014 at 4:16 pm (#)

    […] you heard of evaporative social cooling? It says the people who provide the most value to a social group or organization eventually burn […]

  32. The State of the Internet is Awful - Notes from Mere O says:

    August 27th, 2014 at 5:24 am (#)

    […] you heard of evaporative social cooling? It says the people who provide the most value to a social group or organization eventually burn […]

  33. Dalla veranda alla strada | Bicycle Mind says:

    August 27th, 2014 at 10:56 am (#)

    […] come scrive Frank Chimero, da un portico Twitter è diventato una strada: Have you heard of evaporative social cooling? It says the people who provide the most value to a social group or organization eventually burn […]

  34. 63: For The Articles | Irregularity says:

    September 2nd, 2014 at 10:58 pm (#)

    […] Hang: Social Software Sundays #2 – The Evaporative Cooling Effect […]

  35. Frank Chimero – From the Porch to the Street | Miscellanea says:

    September 17th, 2014 at 9:27 pm (#)

    […] Chimero – From the Porch to the Street Technology Have you heard of evaporative social cooling? It says the people who provide the most value to a social group or organization eventually burn […]

  36. Communities – Planned Obsolescence says:

    December 11th, 2014 at 3:01 pm (#)

    […] active and engaged on Twitter have begun withdrawing. Perhaps the drop-off is part of an inevitable evaporative social cooling effect. Perhaps at some point, Twitter’s bigness crossed a threshold into too-big. Whatever the causes, […]

  37. Guest Post: Digital Publishing – Communities | Medical Heritage Library says:

    January 28th, 2015 at 5:21 am (#)

    […] active and engaged on Twitter have begun withdrawing. Perhaps the drop-off is part of an inevitable evaporative social cooling effect. Perhaps at some point, Twitter’s bigness crossed a threshold into too-big. Whatever the causes, […]

  38. I’m 28, I just quit my tech job, and I never want another job again | Tech Bit says:

    June 30th, 2015 at 6:51 pm (#)

    […] critical mass of engineers, a people we knew and reputable began to deposit apart. The hacker forms evaporated divided to other jobs, mostly for startuppier companies. Many of them had been a fortitude of a […]

  39. dredmorbius comments on “How Not to Sort by Average Rating (2009)” | Exploding Ads says:

    July 9th, 2015 at 12:18 pm (#)

    […] And there are other problems. Limiting individuals to a single vote (or negating the negative effects of vote gaming) is key. Watching the watchmen. Regression toward mean intelligence / content. The “evaporative cooling” effect (…). […]

  40. I’m 28, I just quit my tech job, and I never want another job again | RobyScar - innovations and visions for SME says:

    November 29th, 2015 at 9:13 am (#)

    […] critical mass of engineers, the people I knew and respected began to drift apart. The hacker types evaporated away to other jobs, often for startuppier companies. Many of them had been the backbone of our […]

  41. How to build and maintain communities - @visakanv's blog says:

    June 13th, 2016 at 6:31 am (#)

    […] Read about the evaporative-cooling effect. […]

Leave a Response