This chapter discusses how content influences online
community and how to manipulate content in online community under specific
purposes.

The Whole Earth’Lectronic Link(Well) is taken as the first
example to illustrate the relationship between content and community, where
content should be considered as the most important ingredients in the
development of online communities.

One useful strategy to throw out topics which you expect to
get from your audiences is to start talking about your own story and followed
by others’ responses. I suspect that such kind of strategy is more useful for
the early stage of online community, or under some special purposes, such as
advertisement.

Tips on conducting useful contents are also listed including
“Be personal”, “Be nice”, “Be inclusive” and “Be honest”

Finally, the example about MetaFilter, which is the site
recommending interesting sites
illustrates how Matt Haughey develops this site through conversation
construction, related rules and other approaches. The great success of this
site is not only to provide link to other site, but the original content
generated by the site itself through the comment system.

Both articles this week gave advice and logic on how to start an online community and/or keep a burgeoning one growing. What is interesting, is that both articles seemed to focus on the rhetoric of the language and actions in doing so more than the actual "how-tos" involved in such a process.

For example, Powazek focuses a good deal of his chapter on the proper tone in initial messages to any community, given that follow-up messages will follow suit. He likewise gives advice about knowing one's audience, so that snarky audiences will love snarky messages (20). He then takes this emotional appeal a step further by explaining the need for emotional, personal stories to not only draw people in, but snuff out flames that may occur in more heartless, anonymous posts (21).

Likewise, in the Chen/Konsten/Resnick chapter, they spend a lot of time discussing celebrity endorsements and conscpicious consumption in the latter pages. Again, we're treated to a buffet of appeals that communities can provide to their users. The authors here say to use celebrities to appeal to logic, fanhood, and emotion. As they move on to discuss conspicious consumption by selling faith in a growing product, they're likewise asking community members to appeal to issues of credibility and emotion, so that a user will feel like "if it's advertised on the Superbowl it's gotta be good!" or "if only the best poker players are here then I want to be a part early!"

So my take-aways from both articles is that it's not so much what content you provide, as HOW you provide it. As a result, we're back to the basics of rhetoric online as it is in the "real world": know your audience (even before they exist!) 

Here's at least one data point for your partial chapter: 

> projects_all = read.csv("http://catfish.blockstackers.com/~oostendo/sfdata/200801_sf_projects.csv" , header=T)
> length(projects_all[projects_all$members > 3,'group_id'])/length(projects_all$group_id)

[1] 0.1025422

10.3% of projects have more than 3 members.

--------------------------------------------------------------

oostendo@umich.edu

--------------------------------------------------------------

I think probably the most successful uses of syndicated data for startup are when it can be manipulated, extended, or used in some way which is not available on the host site.

A couple of examples from my experience

- AnnoCPAN importing the POD documentation, normally browsable on CPAN mirrors, but now giving users additional capability of being able to annotate the documents.

- PerlMonks importing all of the standard perl documentation in /usr/doc, but improving it by connecting it to related content on the page and allowing users to add comments to it

- Everything2's import of websters dictionary from the gutenberg project -- which added tens of thousands of nodes and links to the system, some of which were added to and rewritten by users.

--------------------------------------------------------------

oostendo@umich.edu

--------------------------------------------------------------

This is something that I did in my reckless younger days, and probably wouldn't now, but it does seem to work enough that I know it has been tried many times on many sites. When launching a site, part of the problem is that it really can't look "empty" -- you need discussions, small talk, and maybe even a little bit of fighting in discussion forums just to get that comment count above zero and make it look like somebody's home.

The technique I used was basic role playing, where 4 or so people each came up with 5-6 different "users" and registered accounts for them. They had sparse backstories - "The high school student who just started getting into computer programming", "the old codger who has nothing better to do and always talks about ham radio", "the 20-something graphic designer from toronto who uses a picture of his cat as his avatar".

Each person logs in at semi regular intervals (30-60 minutes) with each of their different roles and participates on the site "in character". If you can do it for a few days, it actually looks like a reasonably active community. Don't forget to do a fair amount of bitching as to what the site lacks (but you were planning on adding anyway). Also be sure you quit before you actually publicize the site -- you don't want real people trying to make connections with your shills. Then after launch and (hopefully) takeoff, you can quietly purge your shill accounts and comments.

I really can't endorse or defend this type of behavior -- it's risky because if people find out you can get shredded on Digg, and it's really not honest. But, that said, it is *kinda fun* to go thru your site as "real" users and think up random things for them to say. It really makes sure you use and think about the site the site in a way that your users are going to use it, and it keeps the place from looking like an empty garage when you are trying to publicise it and get it off the ground.

--------------------------------------------------------------

oostendo@umich.edu

--------------------------------------------------------------

I didn't like the example in Paul's paper about how pandora tried to be compatible with last.fm (p.3).

Generally, I would say that a similar user interface is a quite low type of compatibility in any case.

I also think that such "black box search user interfaces" are just very widely used by many search services especially Google (and for good reasons). So I can't imagine that Pandora chose this interface only by looking at last.fm.

I don't want to say it is an totally unfitting example but it was just not very convincing to me. Unfortunately I can't really think of a better example from the top of my head.

Both readings reinforced the idea that "Knowing your audience" is a critical factor in the success of online communities.  One trend I've noticed with the larger e-communities, is that there are key strategic points when communities lose popularity points due to the addition of a feature that does not jive with the ways users interact with a site.  For instance, a huge migration from friendster to myspace took place when friendster added the "see who has checked your profile" feature.  Because these site are by nature voyeuristic to a certain degree, making the profile checking process totally transparent is counter-intuitive.   

Tracking the fall from favor of large social e-communities and / or migrations from one site to another, may provide insight into features that have the potential to turn your audience off in a major way.

Lisa McLaughlin

Having kicked around a "news aggregator for SI" idea for awhile, the Powazek chapter contained some extremely useful information that helped me focus my thoughts and reasons.

The email listserv gives newsworthy items a short shelf life, niche items too much exposure, and pings our attention often at times when we're focused on other things. I'm hesitant to comment on any mail because I don't want to be "that guy" who inconveniences hundreds of people - most of whom frankly aren't interested in the subject or my comment - whith yet another SI-all / masters / open email in their inbox. I suspect I'm not alone.

I want to put together a team to build a "Metafilter for SI" to help people find the best news faster and with less effort, have more control over when they interact with news, and see if people want to discuss things with each other outside of the classroom and all our group project meetings.

This project idea already has a little traction in the social computing club, but would there be enough content? Perhaps - with 300 SI students and an occasional appearance by a professor - if 80% freeride, 15% comment, and 5% contribute daily, there would be 15 stories and 45 comments each day. Sounds manageable and lively enough to me.

Do those figures / expectations sound accurate? Is this a worthwhile project? Would you participate?

The draft was very fun to read since it proposed an interesting aspect of online community: how to start up?

I think no matter how many reasons one can have to contribute/participate, only selfish is the important and long lasting and keeping the site going. Before the critical mass, the site can only provide explicit incentives to encourage people to join: like giving storage space, free email, and even money. After the critical mass, the value from aggregation can compensate for this. As more and more websites improving their service, the efforts before taking-off are getting higher required. Obviously, it is getting harder to start up a site than 10 years ago. Free space and email are not cheap. And the earlier winners have taken much such that it is very hard to be retrievable from online search engine for a new comer site.

So some design suggestion would be: a successful site would serve people well along without any additional social effect, but the social effect would be highly valued in the whole consideration. That means, even you had the idea in mind, for the cost consideration, one can try to implement those immediately needed first.

I ilked the readings this week.

Powazek made a lot of points that seem logical to me: Set the right tone, connect with people on a personal level, seed the community with some initial content.

I also enjoyed the draft chapter of the Online Community handbook. Some of the examples were a very good fit. I wonder if Mahalo would be worth a mention under those communities paying members for contribution.

Under the section header "Building a Community out of Small, Closely-Tied Social
Networks" the second half, on viral marketing and inducing people to invite others, does not seem to seem to fit the heading, imo. It seems that what you're talking about is sites that build upon real life or other existing relationships, but it's not always the case that these real life or existing social networks are small, or that they involve close ties for that matter. The header fits the first half of the section better than the second.

I thought the idea of taking existing community data, such as in the MovieLens case, was really cool. It makes a whole lot of sense to build off of someone else's work, but I'm not sure how often people will have access to such data.

For the Show Growth section, you might mention displaying the membership count, contribution count, or even have contests such as the 2 million book competition on PBS, in which many users guessed the date and time that the 2 millionth book would be entered into the system, with credits as an incentive for winning. I've also enjoyed the newsletters that usually say something about how the site is growing.

I couldn't relate much to the celebrity endorsement section. I don't think I recognize/respect/have an attachment to most celebrities. But if you could count nerdy/techie celebrities such as famous bloggers or the like, I think I might be influenced to try things by them. Or am I? It's always cautious, gotta take it with a grain of salt. I mostly just let it pique my interest... I really thing that the "wanting to be in contact" part of it is slightly farfetched depending on the circumstances. I suppose celebrity endorsements are effective for certain parts of the masses, so it's important to consider.

I liked how both of these readings addressed that content is really important to starting any community. Powazek recognized that communities start with content. And in the Chen, Konstan, Resnick paper, a lot of the discussion on bootstrapping involved getting the content for the community going.

From my own experience, I know that I don't go hunting online for more people to become friends with. I am looking for content - I want recommendations for good books to read, I need help programming a web crawler, I'm trying to figure out where I can find a good game of Ultimate in the area. The sites that meet my information needs are the sites that I will keep going back to. If the site provides that content better via community features, that's great. I may even get involved in the community. But I'm really there for the content.

I really like the paper, and I think it will be a great resource.  A few little niggles/suggestions -

I think that the New York Times example, page 4, is problematic.  I think that the point is good, but it is the wrong example.  The Times always had critical mass as a paper as did its website.  Plus, it is a very special, unusual franchise.  I would argue that the Times absolutely could have launched a news discussion site from scratch leveraging their brand.  It was a strategy issue, an issue of their role as journalists, a brand issue, etc. that made them reluctant for many, many years (approx. 10 years) to take that route.  

http://money.cnn.com/magazines/business2/business2_archive/2006/09/01/83...
 
Nisenholtz, 51, dates the turning point to early 2005, when, as he explains it, "the executive committee of the company created a strategy that said we're in the business of convening communities here; we're not just in the business of pushing information at people."

In the section on Building Community from Small, Closely-Tied Social Networks --

Doostang would be a good example for referral incentives.  By getting contacts to register, users have access to some additional privileges / job-listings.  

Also, what Linked-In is doing with the option to use a few popular services to grab social graph data is becoming pretty standard - I see that approach being used in many places.

Expectations:  Look/Feel like you're already successful --

We used B.J. Fogg's web credibility research to do web credibility assessments for info. architecture class.  My project is a very Web 2.0 vertical search engine which had too many new functions to fit well into the c. 2001 framework, but the principles were useful.  
Exclusive Communities Trend --

It might be worth mentioning the trend towards more exclusive online communities and filtering mechanisms - A Small World is a well known example, and several sites have big TV marketing campaign that emphasize the advantages of exclusivity for one reason or another (theladders.com, eharmony).  From a business perspective, it is very smart to aggregate affluent, niche audiences.  Businesses have done this successfully in print for many years, and I would expect this trend to grow.

Some general thoughts about the Chen, Konstan, Resnick article: 

I agree with Liz that celebrity endorsements can sometimes sway opinions, but this section of the paper seems to be more of a business tip on how to do marketing than one specific to online communities. Some of these examples, like Sam Waterson/TD Waterhouse, aren't related to online communities at all. Plus - most communities probably don't have the resources to consider hiring a celebrity for an endorsement...what are they supposed to do? In my opinion, it seems more useful to focus on developing relevant content and features for a community than on something more superficial like an endorsement.

Other points in this article I definitely agreed with, like the benefits of starting small and focused and then building a community. A good example of this is how Facebook started - by first being available only to Harvard, then to the other Ivies, then all colleges, then high schools, then everybody. If it had started with being open to everybody, I don't think I would have initially joined because there wouldn't have a been a critical mass of my friends joining as well. 

This paper also had a number of good points about how to make people become the early adopters of your site, at the point when you don't have a lot to offer them yet. Still, I think even if the community is not there yet, there has to still be some benefit to be gained from the site, and this benefit comes from having good content. As Powazek said, content comes first. You have to have a reason for them to come to the site that will get them excited and wanting to tell their friends. The networking externalities theory said this too: "if no one believes the site will succeed in attracting users, no one will join". The main reason I usually think a site won't succed is because of lacking good content.

One last rant (for the semester!)...I liked the line in Powazek's article saying how web surfers have come to expect community tools on websites. "These tools are now so ubiquitous" he says, "they've nearly lost all meaning."

I can remember what he was talking about, in the context of the web in 2000. New services were popping up every day to embed polls, message boards, and more into your website. Everyone seemed to have a guestbook for users to leave comments. That was then, but I think Powazek's point is even more true now with Web 2.0 and all the ballyhoo about user-generated content. Sure, there have been many successful websites that have benefited from this (Amazon, delicious, etc.), but on the whole I think there's gotten to be too much of an expectation that if you just build a website with community features, that users will come and you will be successful. Browsing through some of the half-baked startup company ideas on TechCrunch proves this. :) I think the Chen, Konstan, Resnick article is exactly the type of thing that entreprenuers ought to be reading to remember there is a right way and a wrong way to go about building a community! 

I guess I'll write about how I bootstrapped Anime Nano in here, since I dunno how many people actually read blogs.

Anime Nano was started when I saw a need to create a balance between Animeblogger.net Antenna and Blogsuki (as I described in class last week). There was some blogfighting erupting over the whole "walled garden" vs. "crowded commons" and Anime Nano was supposed to be a combination of the two. Here's the initial post I wrote about Anime Nano.

Anime Nano gained a critical mass early because it leveraged a few things:

• A need in the already existing anime blogging community for a custom solution to the quantity/quality problem
• Existing "loyalty" from a fairly well known anime blogger (I endorsed my own product, obviously)
• "Viral" propagation due to the inherent properties of the community (anime bloggers wrote about anime nano, leading other bloggers and readers to start using it)

Anime Nano was basically a content aggregator, so the task of getting content for users was quite easy. The issue with network effects wasn't present because it reached critical mass within a number of days. Overall, the site was a success very quickly. The community really accepted it wholeheartedly. It still gives me fuzzy feelings when I look back at old blog posts about Anime Nano.

In the early stages, forum use was very active. As time passed, people have been using the community features less and less.

I'm sort of in a state of flux now with Anime Nano. Taking this course and having these online discussions has helped quite a bit in what I think will be my roadmap for improving the site come this Summer when I have a bit more time for coding.

I've already developed a quick "level" system for the forums, which was noticed quickly and led to the most activity the site has seen in quite some time.

As my last post here (Except for my blog, I guess), I'd like to say thanks to all my classmates and Paul for the conversation. If you have any other comments for the site, I'd love to hear them directly (hktruong@umich.edu). Anyone who emails me gets points! 

John Blair 

As I read these articles, a consistent thought kept occuring to me.  All the actions / theories described in both articles are merely means to increase the odds of success, but they can not guarantee a new site / community will be successful.  The definition of success I realize is flexbile and subjective.  I think the ending to the Powazek article summed it up very well from the initiator's (community builder) perspective at least.  "pick something your passionette about, devote the necessary time - for as long as it takes and be open with your users".

Given that nothing is guaranteed, how can the odds be improved beyond what the methodoliges and theories stated by Powazek and Chen, Konstan and Resnick?  As a variation on the idea proposed in Chen, Konstan and Resnick on page 9, that "sites start with a scope narrower than their eventual aspirations", I wonder if it's possible to more accurately assess what critical mass is and then design for it.  This was somewhat implied on page 2 of the Chen, Konstan Resnick article in the statement that "it's important to influence potential user's expectations in order to achieve the "good equilibrum"". 

Suppose you had to build a business case for your new online community and the banker who may lend you the money wanted to know your plan for success, including how long it would take and how you would know when you reached it (critical mass).  Since the bank is obviously interested in managing their risk, these are very valid questions from their point of view and their decision is based on how well and objectively you present your business case/plan.  How are you going to know when critical mass has been achieved and how can you demonstrate this objectively to the banker?

Referring back to the concept of starting with a smaller scope initiially, you could possibly develop a "test" site (which in all liklihood would be the early version of the real site) and obtain metrics like number of members, content volume, user feedback, member/user areas of interest / postings, growth of content, focus of content, member reactions to site changes, administration of the site, number of admins required to administer/maintain, etc.  Answers to these types of questions may further improve the odds of success, since they can demonstrate to a certain degree how you, as the site admin/creator are going to design into your site the ability to know when critical mass has been acheived so that the site runs at an optimum level without severing the degradation or growing pains suffered by many sites that grow too large and don't have a plan to address the volume of growth.

These two pieces are really interesting to read. I learn that a website always has to remember that content comes first. However, it’s especially hard for a starup online community to gather many participants to generate enough and desirable content to attract more users. These two readings, especially Chen, Konstan, Resnick one, give me great ideas about what strategies I can use to make the online community take off successfully. However, some questions come to my mind, while I read the Chen, Konstan, Resnick piece,

• Syndicated data can help a new site to get sources, and MovieLens is a great example. But I wonder if it’s easy to import source from an existing online community, especially if the existing site still want to operate continuously. This is because the existing one may not want to help an initial small community to become a real competitor.

• Conspicuous consumption can signal users that the site owner has a commitment and believe he/she will eventually successful, because conspicuous consumption can differentiate low type and high type automatically. But I am not sure if all site owners really know what type they belong to. For example, is it possible that a site owner thought he is a high type and invest a lot of money, and then he realizes he is a low type since it doesn’t work.

Also, I am not sure how to estimate the critical mass that an initial site needs to take off.

ubuntuforums.org site was started one month after the Ubuntu project was announced by Mark Shuttleworth. It was started as a mirror for the mailing list. (Based on my inference fromt the forum archives). I don't it was an automated mirror or the ubuntuforums.org owner manually mirrored the content. Irrespective of the approach, site capitalized on content generated in a different medium.

I think the early mover advantage helped ubuntuforums.org quickly ascend in popularity. The site was started shortly after the Ubuntu project was announced. So all early users of Ubuntu would have known the site. So it is not surprising that Canonical the commercial company supporting Ubuntu decided to adopt ubuntuforums.org as the official support forum instead of creating another online help community.

One design constraint that seems implicit but not directly discussed in this week's readings is whether you are building an new online space for an existing community (say, the alumni of a school, or SI students, or even the Indian expats in Minneaoplis) vs. trying to create a new community. It seems that you'll have different options in each situation. If you're trying to create the content for an online space and you have an existing community, you can have a work session where the existing community helps populate the new space with content, and maybe works out technical glitches.

There are probably some other good examples of how the startup challenge is different (I'd almost frame it as an adoption problem (existing community, new space) vs. a recruitment problem (new community)), but for whatever reason my mind is drawing a blank right now. Thoughts?

Another point that I think is implicit in these readings but that could stand to be drawn out is the power of the personal invite. This appears somewhat in the community handbook section on "building a community out of small, closely-tied social networks," but this seems to mostly focus on leveraging users' existing connections. That works well in some situations, but I think there are also situations in which this is not required to give a personal invitation or something that makes the user feel special. 

One example is betas. You let people indicate their interest (ok, we're back to the recruitment issue for this, or providing some expected benefit, but u_unsocial includes the added utility of being "first" or "exclusive" if you get to participate), and then invite some in to try out your tool.  I don't know of any research that supports this (anyone else?), but it seems that this will create initial users who are more committed than an open beta might.

You don't even have to let users indicate their interest, you can just approach some people that you think might be interested (but you have to have this list from somewhere... which is another challenge) and invite them to try it. This also may give you the benefit of working with a small initial group to set the tone. This combines the benefits of a personal invite with some of the ideas in the "early commitment approaches" section (status for being early).

To start a new online community, one has to integrate the community features into your website. You want to leverage all your existing traffic to best effect - keep everything on the site that people know; i.e. the one that already exists! Web users are lazy. Keep everything they need on one site.You need to have a very specific target niche. On the one hand, this can be beneficial as you will naturally have less competition. On the other hand, you may struggle to get the right visitors to your website. For a forum to be successful, it needs content, traffic, members and interaction.Also you need to have a targeted niche. Take a forum site for example; I would guess it might be better to start off limiting the community aspect of the site to a blog. Blogs take less work to turn into a success, yet they can create great communities through the commenting system. Once have built a strong community around the blog and have strong repeat traffic and loyal members, it will be in a better position to launch a forum and ensure it is successful. A forum is a whole new kettle of fish and is a big step up.Plan for a forum, but consider holding off until you have the traffic and time to ensure it is a success. Nothing turns visitors and potential members off more than an empty forum. Whatever developing any online community from scratch is hard work.  

It is important to identify when to expand the community. For example: Flickr was hugely popular and it was expected to include video sharing for a long time but they did not do that. youtube came and became the champion of video sharing which Flickr could have achieved based on its already existing userbase. (similar to what MetaFilter did when they lauched Ask Metafilter - a question-answer website in 2006)

Developing communities require great marketing skills that include
identifying the userbase, targeting them precisely, listening to their
voice (aspirations), understanding them, and finally..delivering on what
they want. Consumer Internet is a tricky business: Because of the viral marketing, a good site could get popularity very fast, similarly a bad site could be trashed equally fast (maybe even faster).

Succesful online communities have understood their
customers well. For example: Digg: It started with very targeted focus (on
technology) and expanded to other domains. I agree with the articles
(both), that early stage startups face scarcity of resources and
therefore focusing on targeted niche is a good idea.

Blogs have become quite sucesful in building a community like feeling around the "main blogger". Twitter, Social Networks, or Blogs could be used to create a small userbase which could be moved on to the new community website. One of my friends on facebook created a "facebook group" and invited a lot of his friends (and their friends) before launching the facebook application.

From last class: Archiving the data of the old communities: If there is such a service available in the marketplace then it could be used to provide data to bootstrap new communities in the same space.

I'm poking around on the internet while I finish my last paper for this class, and I came across an example of something Amazon did early on in
bootstrapping their book reviews - offering $1000 worth of books in a
"Book Recommendation Contest."  It's in this book from the 90's, Net
Gain: Expanding Markets Through Virtual Communities, which I'm finding
to be quite interesting (from what I can read via Google Book Search). 

It might be interesting for you to talk about this sort of contest as
another method of bootstrapping via selfish activities producing social
content.  You did mention payment, but guaranteed payment and
competitive prizes are different types of motivators.  Anyhow, maybe
the book, though old, is worth a peek, if you haven't already.