This point has not been lost on the web. Nothing prevents a user from going to a different website if they want to. Often, the execution of this is mere seconds. Thus, the SwitchingCosts are extremely low. Just click. This puts a severe limitation on the behaviour of the host site as its users may just abandon it if it doesn't satisfy them. Thus, obnoxious GodKings may be avoided, as well as boring ghostsites.
Many commercial sites have sought to increase the switching costs by doing a number of acrobatics. One strange practice was an almost CargoCult practice of forcing users to register. The belief was that since registration was so painful, users would not want to re-register at the competition's site. The pain was indeed so great, though, that many users just don't bother using registration-required sites at all.
Community-oriented sites implicitly increase their switching costs by relying on emotional attachment (cf. RewardReputation). If a person has invested time in, made friends on, staked out a position to a site that person will have created an emotional bond with that site. This will cause the person to stick around through thinner times whereas with another site in similar circumstances the person wouldn't even give it the time of day.
There is a major caveat. Emotions are powerful and relying on them puts them under a lot of strain. If your users become agitated, it's better if they leave; but if they are bonded, they may strike out in anger. This causes FlameWars at best and potential site attacks at worst. And we all know that the people who know the site are the ones who exact the most damage. See CategoryConflict to see how emotional attachment can cause grief.
A limitation on RightToLeave both online and in MeatSpace is the small and finite number of relevant communities. The best example is geographically isolated locales. In a town of 400, there all too frequently is only one bar, only one church, only one bank. People who spend a lifetime in such places have highly developed skills and coping mechanisms for dealing with old baggage. They find a way to take their new girlfriend out dancing at the bar where their ex-wife works. ForgiveAndForget becomes an art form (The dark side of this is a decreased willingness to take risks). The situation changes somewhat in more urban areas, where it is possible to change venues and jettison essentially all unwanted baggage. This is true even for economically encumbered relationships, like jobs, where though the SwitchingCost is considerable it is still possible to find other employers.
But even in urban areas, as the desire for interaction grows more specialized, choices become limited -- much like the small town. Even in a large city there may well be only one Morris Dance team in town, so the RightToLeave is contingent upon willingness to leave behind the activity as well as the people.
The same is true of online communities because of MetcalfesLaw. While there are many collaborative encyclopedia projects, one clearly has an order of magnitude more success than the others. There are only two forums for the discussion of intensive grazing; in like fashion there are a whole host of specialized subjects where the number of relevant communities is starkly limited. However, it is so much easier to create your own online community than creating your own bar that at a certain level this obstacle is not restrictive. That is particularly true of WebLogs and MailingLists where the power to link is nearly equivalent to the ability to participate; this is less true of more social spaces like MultiUserDungeons and wikis which have accrued a lot of social capital that is non-reproducible.
Another major caveat lurks. If as the site proprietor you seek fame and glory by maximizing your userbase, you may have to make compromises in your vision (cf. RadicalInclusiveness). Populism and demagoguery are the tools of any successful lizard. On the other hand, you can immunize yourself against people exercising their RightToLeave if you aren't overly concerned with the number of site visitors. After all, it's legitimate to come to the understanding that this place needn't be all places. In that respect the RightToLeave is actually beneficial. You can take advantage of it by suggesting that the Internet is a large place built out of small niches. If this niche isn't what you want, there are countless others. There's no sense growing agitated about it. After all, the SwitchingCosts are zilch.
A not-necessarily-emotional mechanism works against exercise of the RightToLeave: Participants in a community or project may have invested more than just emotional energy, but concrete time or other resources in a site (cf. VestedContributor, CommunityMember). The first stage of this investment, the time that person has spent familiarizing himself with the project and its resources (reading posts, reading code, whatever) can often remain all but invisible to the other members of the community as it does not offer any immediately visible or verifiable effect. Further, any individual will then recognize that others have also made similar significant investments in that project as well, and may be loathe to repeat their efforts in another venture with the same, or similar goals. So, two strong disincentives work against the tendency to just walk away: Valuing one's own investment, and valuing that of others. These complex situations may demand legal structures above mere volunteerism, such as a NonProfit corporation.
Nonetheless, the RightToLeave should not be underestimated. It's important to understand when creating a new project that unless you have an intrinsic pull to trap members in participating, your project will fail if your members exercise their RightToLeave. This concept has mostly been lost on people so far, causing site proprietors to exercise their own powers to abuse users. The example of the registration process above is one example, SlashDot is another example. In both situations, users have left en masse--eventually (let us still not discount the emotional SwitchingCosts too much). Things are harder if you have no intrinsic pull, if your project is only a social construction like a new Internet standard. Then you're even more at the mercy of your own membership because they can impeach you simply by ignoring you. Your power is only in their hands--in their heads--and thus it is really their power on loan to you.
A particular corollary to this realization is that the RightToFork is not demanded de facto or even desirable. ForkingOfOnlineCommunities is a disruptive act, possibly destructive; however, the belief that the leader of the project acts not only in good faith (AssumeGoodFaith!) but also in selfish interest to maintain her emotional stake gives strong motivation to keep the peace with her own supporters enough to discourage secession. After all, a fork of an online community is really people exercising their RightToLeave whilst maintaining their emotional stake.
The Case of the Blackhole of Love
An example of an individual too emotionally attached to leave:
The individual was blackholed after supporting the infamous "Troll Investigation" thread that scored thousands of moderations after the editors continuously moderated it down. For his reward, he states, "CmdrTaco told me in his actions to fuck off," but instead of leaving he is "protesting against CmdrTaco by not subscribing and blocking ads." He justifies staying by stating that he likes the "majority of the community," and he claims he'd subscribe after CmdrTaco treats him fairly (FairProcess) and he receives an apology (ForgiveAndForget?).
It is interesting to note the use of "too emotionally attached" (and previous references to "emotion" above) as a pejorative. One wonders if all online interaction is to be all Mind and no Heart. Whether caring is considered a liability. Whether the author of the phrase fancies himself a Vulcan.
Certainly, emotion can promote some tunnel vision--in these cases, sometimes humor can restore perspective. But one need be very careful not to tell people, effectively, that they shouldn't care. Such dismissal of people's reactions to their time and interests can be counterproductive, leading to an escalation of conflict, rather than a restoration of dynamic balance. --anon.
Devaluing emotion wasn't the intent of the above. In fact, it was the absolute opposite. The role of emotion in an OnlineCommunity is central, clearly, as without it no community could form. Thus all leaders must understand the value that an emotional stake has in forming their communities. In the case of the Slashdot poster, he was "too emotional" because he should have just left. As pure unabashed infotainment, Slashdot isn't worth "saving", whatever that could mean, since the purpose of Slashdot isn't to be a liberalist discussion forum. You can't bend a rock. -- SunirShah
So, it would seem that the more important component of your analysis is that the person's mistake was in misperceiving the nature of the site, rather than in caring too much about it. --anon.(ibid)
No, the mistake was still caring too much about Slashdot, a site with "junk food" emotional value. Even if the individual failed to recognize the malleability of Slashdot, the question still remains why the individual bothered in the first place? -- SunirShah
In hindsight, one can ask why someone bothered in the first place, and it seems like a reasonable question. But one "bothers in the first place" gradually, incrementally. One participates at first just a little, perhaps just reading or lurking for quite some time, all the while building up something of an investment in the place. By the time a "fight or flight" (RightToFork/RightToLeave) situation comes up, the person is already invested, whatever someone else's subjective estimation of the value of that investment.--anon. (ibid)
While undoubtedly people will still remain emotionally committed to things they shouldn't forever and ever, the prescriptive sentiment on this page is to warn people to provide themselves perspective. That requires experience and maturity. Failures happen, responses to which are covered elsewhere. -- SunirShah
Contrast this with situations where one has no RightToLeave, like a WorkplaceWiki?. Do the additional MeatSpace incentives/rules/whatever act as sufficient BalancingForces?
Studies to date have shown that information technology in the workplace is a common weapon of choice for those seeking additional power, following Foucault's analysis of information as power. The RightToLeave functions most strongly in a doubly open-ended volunteer situation, where the intended goal is non-critical. This of course describes the least important of all organizations. When additional forces are brought to bear, simple belonging and leaving are no longer sufficient indicators of value as other motivations may interfere with these processes.
See also WikiSubcommunity