Social Software and campaigns
Many of my friends are Howard Dean supporters, others work for one of the other Democratic candidates. A few friends are republicans, a few are Libertarians, others are even anarchists (mostly anarco-capitalists as complicated as that can be).
As I watch the current Democratic primaries I find myself pulled in many directions and have not yet found a candidate that I fully support. At the same time, as a person who is very interested in the Internet, in software, and in the power of networks (indeed I am forming a group focused on Network Theory and plan to hold a conference on the subject in the future) I have been watching what the candidates do with the Internet to organize. In particular I have been following what the Howard Dean campaign does, though not exclusively and have not had the time to get into it as fully as I would like to.
After the first two states, the primaries are shaping up to be very different than in past years, but not, perhaps, as different as they were expected to be. Specifically, technically the campaign is still anyone's game - only a very small percentage of the delegates have been selected and while there are some fairly clear losers, it is by no means certain that Kerry will emerge the winner, or that Dean the loser.
That said, just this afternoon I recieved an email from a friend of mine here in Chicago who has been a staunch Dean supporter for quite some time - periodically mailing to his friends around the country about his support, even in the face of various public gaffs and blips. His most recent email, however, stated that he, along with many other Dean supporters he knows, has decided to shift towards supporting Kerry.
His problem with Dean is not a specific policy or issue, but rather Dean's seeming indifference to and lack of ability to present and project himself as a presidential candidate That is, while perhaps it should not matter, he does find that appearance and perception is vital to being able to win - not just in the primary, but in the general election.
In thinking about Social Software and the Dean campaign (as well as other campaigns) a few thoughts occur to me. For one, time is a vital element to all network activity of a campaign, unlike in more general cases such as the dating networks of Match.com or Friendster, a campaign has to not just create and mobilize and network, but focus that activity around very time specific requirements, a primary in each state and the the general election. For another, the network does not exist in isolation, there are many competing networks (other candidates, the "other" party) as well as many diverse forces that will react to the network's actions. As well, an open network is just that, open, you have to assume that not only are all of the actions public, but even the process of chosing and selecting those actions may very likely be observed (and perhaps even influenced) by people active in other networks.
This final point bears some further discussion. By this I do not just mean the "traditional" model of espionage between campaigns, I also that open campaigns mean that people who are making up their minds may still be active in and influencing many different networks. Consider as well that opening up the process of decisionmaking as well as the final decisions means that the messy process of reaching consensus (or at least a decision) likely plays out in the public eye, observed not just by supporters, but by potential supporters and competitors alike. Humans being who we are, it is also to be expected that gaffs will be focused on over the positives and that people will quickly leap to conclusions, even if a decision has not actually be made.
In any network it is very easy to assume that everyone is alike. At a very least most networks (and especially online communities) assume that the participants each have a set of things in common, over time these grow to include generally a specialized vocabulary and a shared worldview. The danger that this holds can be seen as you observe how networks change and grow over time, the more complex the shared history, the harder it can be for new members to join and once joined, it is that much harder for the new members to influence and shape the network. Consider a startup, in the early years every employee has a critical and vital role in the company (if they don't generally the pressures of a startup either make them have a critical role, or mean they have to find a new job elsewhere). However, as the company grows and the roles become more standardized and formal, the influence any new employee has over the course of the company diminishes. Until as somepoint a company becomes a big company full of mostly employees disconnected from driving the future of the firm and seemingly unable to shape it in a meaningful way (or generally to benefit from such changes either).
In a campaign, the early stages access is generally open and easy, most staff are unpaid volunteers driven by a common set of beliefs and vision for the future, and any given person's role is not hard to see and feel.
However, if the campaign is to move beyond the formative stages and to wins in primaries, it most move beyond just activists and include regular voters in numbers that are larger than the opponents. Not only must the campaign motivate, identify, and draw upon this support, it must begin the process of building to handle every greater size and scale, eventually enough to win in a national general election.
One aspect of this, I conject, is to build support not just at a national level with nationally known leaders and among the small "activist" population around the country, but to build support from local candidates and officeholders throughout the country. These local leaders, in turn, provide a connection to the non-activist majority of the country. "All politics is local" is a famous quote (source I don't know but suspect Google would have an answer).
As such, "social software" has to do more than just provide the means for activists to talk to activists, it also has to provide access to the campaign by the far larger majority of non-activists. Further, the software has to be careful that it does not mistake the voices of an active minority for the views of the majority. The danger that many commentators have cited about the Dean campaign's focus on Iraq as a rallying issue, is that for many voters Iraq is not the primary issue of importance, but it could (as it is for me) serve as a negative point - i.e. much as I would not vote for a pro-life candidate, I am reluctant to vote for a stridently anti-war candidate because it does not give me confidence in that candidate's leadership in future crises or non-black and white foriegn policy problems - i.e. other cases where the US's "National Interest" might be opaque, groups such as the EU or the UN reluctant to engage, but a major and serious crises is unfolding that the US could, if we chose to, engage, stop and prevent (Rwanda comes immediately to mind as a low point for both the US and the UN).
So the challenge is to engage people to contribute - both money, time, and ideas. Without losing sight of the difficulty in engaging the non-activists and in getting their voices and opinions into the mix.
Like any software, the majority of users of a piece of "Social Software" will not modify it or customize it from how it is presented to them. So, in the case of Social Software, vastly more people will look at the front pages of a site than will customize it into a "my site" view, or who will go through a complex registration process, or who will proactively search out communities of interest. Further while the ratios can vary, in most online (and indeed offline) activities a small percentage (sometimes as low as 1 in 100) of the members are those who participate - by posting/emailing online, writing comments, or talking up offline. The majority of people will read or listen without speaking up.
One thing that groups such as MoveOn.org have gotten pretty good at is helping reduce this burden, in no small part by doing a lot of work up front. Thus, instead of sending out a mail such as "call your representatives and senators", they will send out a highly customized mail with detailed instructions on how to specifically connect with the specific representative or senator for the person to whom the mail was sent. Or instead of asking for "donate money" they ask for something more specific "give $x for this specific activity" and then also provide a wealth of immediate feedback on that giving - both showing what it is doing and showing how much as been raised.
But writing, talking, even contributing money, is a different act than voting. Voting is irrevocable and is, most of the time at least, a representation of making a decision - this candidate or the other (at least here in the US we don't generally have vote off style elections where you indicate a level of support and second/third choices etc, our voting tends to be either/or votes). To get people to vote requires a number of specific and each slightly difficult steps.
First, a person has to be legally registered to vote. The process and complexity of this differs from state to state. In some states this also requires a registration for a given party (states with "closed primaries", in many states however primaries are open so independants as well as declared members of a party can vote)
Second, a person has be aware of the date of the local election. Especially critical if the process of voting absentee is difficult, but even if it is easy, the steps to vote in that way must be known and followed up on. This awareness of the date has to penetrate into a freeing up of the time to get to the polls to vote, as well as time ahead of voting to become aware of the candidates being presented (especially critical if there are multiple elections being held at once, all too often people won't vote unless there is at least one election where they are personally involved with the outcome - typically of a very local race)
Third, a person knowing the date and registered to vote (or aware that they can register at the polling place in the states that allow that) has to know where their polling place is, and take whatever steps are needed to get there while the polls are open. Frequently this means taking some time off of work, or at least leaving work early or arriving slightly late. Why the US seems to generally hold elections during work days has never been clear to me, but I suspect in part it is to keep turnout somewhat down.
Fourth, a person knowing all of the above, has to actually take the time to get to the polling place and vote.
Here social networking software has to tow a careful line. On the one hand, peer pressure is a very, very powerful thing. As I cited above, all politics is local. While we talk in broad generalities, at the end of the day we tend to act in ways that are in keeping with our own immediate circle (which in the age of the Internet might not be people we live with or near). A committment to a distant, remote group is one thing, a committment to a friend is a very different thing entirely. However, voting is also, at least in non-caucus states, a very private, personal matter.
Additionally, social networking software has to watch out for the "group think" problem that any crowd - whether virtual or "real" can fall under. This is the phenomenon that most people will generally agree with a crowd around them, whether or not they privately (think when they walk into the voting booth) agree. This problem is a serious one, but also one that I think could be helped by social software, however the challenge is time shifting of participation. Few people will publically (on or off line) speak up with a view that is counter to how they percieve the group views things. Generally they will either not talk, talk in vague terms, or literally agree with the group (at least their perception of the group) while privately disagreeing.
Writ large, I wonder if this is some of the challege that the Dean campaign is facing. A lot of people finding some common ground, but privately holding back on many opinions that they feel are unpopular (at least with this group). Over time, these private reservations may grow and when faced with a private opportunity to act (i.e. vote) people appear to be acting in ways other than how they may have publically spoken up.
Consider the issue of Iraq. Within Dean supporters I suspect there are few who feel comfortable publically saying "I am, was, and will be for the war".
In corporate meetings a technique can be used to great effect. Rather than asking for public participation and feedback at all times - leading most to speak in a way designed to impress (and echo back) their boss; private feedback channels can be established - such as a private vote, poll or suggestion box - typically delivered electonically, but can be as simple as writing notes on a slip of paper. The key, is that people should all give this feedback, before hearing the results. As well, after it has been shown that there is a diversity of opinions and views, more feedback should be solicitated (as many people the first few times around will still think that they are being watched/judged on the basis of the feedback.
Often in such meetings it is very helpful to have people publically seen as neutral involved - whether people from another part of the same company, or frequently outside consultants, the presense of a third party often prompts people to open up in a ways that within a group they might not.
For social software, I think this is where the line between openness and privacy has to be carefully watched - what people will say in "public" and what they say and do in private are rarely completely the same. If a campaign wants to get a sense not just of what people say at a meetup, rally, or online petition, but what they do when entering a voting booth, the campaign has to provide opportunities for private feedback as well as public. Further a campaign has to highlight and emphasize the range and diversity of opinions if the base of supporters is to be grown beyond some core set.