a continued discussion on project DAO


As discussed in the project DAO, I had a lot of dissatisfaction from the results and insights gathered from that experiement. Therefore, I want to use my PAX final project to further study the reasonings behind online community emergence. I value this to a great extent because I see community building and operation as one of the most important skills to have in the next decade. For further explanation please refer back to the origianl DAO project page. 

ideal system:

The ideal system for online communities is to view/design the social aspect of a community model as the equivalence of its physical counterpart. We note that the community model in cities, villages, tribes, etc. are all self sustaining, clearing out the old and make way for the new. However, in the case of high growth online communities, we face the unique challenge of morphing the community model from tribes to cities in very little time. Moreover, the emergence of communities are often overlooked as they are such case by case scenarios. Therefore, in order to understand how communities scale we’d want to study how they initally form. This is due to the fact that scaled communities are essentially supersets of smaller communities, which means we need to understand how they initally form. 

project setup / overview:

The very first step into achieving the ideal system should be finding a way to better understand how communities form. One of the core issue is that communities form in a very unpredictable way. Therefore, finding a fitting environment for the community to emerge can be extremely difficult. As a result, I initally prepared two different setups - a twitter bot that tweets out interesting creators and posting daily survey in the community I tested project DAO in. However, I did not gain traction as it is hard for people to discover bots and reply to surveys with no with the post. This ultimately attracted more attentions, and we ultimately asked those who are interested about their backgrounds and perceptions on community to extract insights.

key questions / answers:

( how do we create the emergence of a community?  
-> NEED to piggy back off of existing active community
-> NEED to show potential members interesting topics

( how do we extract insights from interested parties?
   the core question to this project
-> NEED to conduct interviews 
-> NEED to gather qualitative data
-> NEED to extract higher level observations and form patterns

( how do we attract enough people?
   assuming they were not aware of the presence of this community
 -> NEED to publish this in a place that is visible to the target audience

( how do we create engagement?
   to create tractions with content instead of other incentives
-> NEED to create interesting post for people to engage in the first place

experiment setup V0:

I first setup a twitter bot that ran for 3 days straight, posting content of interesting artist in the morning and evening in est everyday. After 6 posts, I realized that I do not get any tractions despite using popular hastags or taggin related accounts. I realized that this would extend way out of my deadline if I were to even have any traction to test on.
I also made a survey in hope my exisiting community members would want to react and particpate in. However, after posting the survey in the community for 3 days straight also twice a day, I only got 2 responses from two people I already knew personally. I realized that people are less inclined to participate when there is no incentive for them. More importantly, the fact that there is no interesting content made this even worse.

This made me quickly change my plan and combine the twitter bot idea with the community idea in hope to gain more tractions for building the emergence of this community. 

experiment setup V1:

I combined the twitter bot idea with the idea of piggy backing off of existing community by posting the post I originally intended for the twitter audience to my discord server. This solved the issue on both sides. On the twitter side, I needed more audience and had a hard time since people were not seeing my content. On the discord side, people needed to see the interesting content first before they can react to the survey since there was no interesting content to begin with.
As a result of this change, I did get more responses from the community since I gained both exposure and also had interesting content to attract the right people. I posted artists in the early 1900s bauhaus era such as Paul Klee and Josef Albers in an attempt to attract the attention of people interested in that particular era of art. If they become interested they will be led to a survey that asks them about questions regarding their decisions to click on this as well as asking for consent to do a further interview.


testing session description:

From the experiment, I could see that 56 people clicked on the form and were interested. 23 people started the form and answered some of the questions. 7 people finished the form and 3 said yes to a brief interview. I tried contacting the other 4 people that have answered the question but did not say yes to doing an interview, and 2 said yes. Ultimately I had 4 participants for the interview with me as one of the other person dropped out due to scheduling conflict. The interview took a more free flowing method, with me guiding the participants into the following questions and letting them elaborate and extend the questions from there on.
    1. What made you engage in a community.
    2. What prompts you to choose such community in the first place.
    3. What’s the relationship between the size of the community and your personal engagment.
    4. Elaborate further on how your exisiting communities emerged.
    5. Why did you choose to engage in this particular community/post.


-> Case 1: Belonging (positive relationship with more identities
Identity plays an important role in the engagement of a community. People often come together because they share the same identity over at least one topic. During the discussion, I realized that people tend to engage more in a community if they share more than one identities. Looking at our friends and communities, this is often true. We share many more identities with our friends that we spend more time together and internet strangers. Moreover, the childcare and skincare section of our existing community has more engagement and people are more willing to engage in it because they already share the identity that got them in here in the first place.

-> Case 2: Purpose (positive relationship with stronger purpose
Purpose is vital to the engagement of a community due to the fact that it is what bands people together in the first place. During the discussion, I realized that people tend to engage more in a community if they feel it has a stronger sense of purpose to them. Moreover, the engagement would also see an increase when the purpose is more aligned, unifying the community. This explains why passionate sports fan and financial related community usually has a high sense of engagement due to the fact that they either feel very passionate about the topic or the topic is very important to them. Moreover, this would explain why our physical community disbands after the group moves away since the purpose of being physically connected is less strong and unified.

-> Case 3: Behavior (positive relationship with similar behavior
Although this is not as talked about as previous points, two people did point out the fact that behavior and the tolerance of other members’ behavior is a very important factor in keeping a community engaged. This intuitively make sense as we tend to spend time with people that behave the way we do or like. Although this does not weaken the overall engagement of a community, behaviors of certain members (especially vocal minority) would keep individuals away.

quantitative result / insights:

All in all, the result from the interview was intuitive but not initially obvious. Moreover, the methods of testing experiments and creating a small scaled community scaling was incredibly difficult to achieve in the first place. I realized that piggy backing off of initial communities or using a platform that has engagement in the first place can be extremely helpful to ultimately achieving the desired result. From the data, we could see that out of the 1000+ members in the community, only 50 saw the post and initially engaged with it. Moreover, only 2 people initially agreed to participate while 7 people finished the form to show interest. However, I already think this a very optimal outcome as it is extremely difficult in this day and age with a lot of information overflow.


The overall testing showed that there are things we can manually curate in order to get a strongly bonded community or create the emergence of a community in the first place. We see that community bond strongly or emerge in the first place when the follow the three criterias we have observed: community purpose, members’ identity/interest, and behaviors/culture. This is in reality extremely helpful because it can help us to identify and form tighter communities. Moreover, it can better explain the emergence of communities.

Thank you for reading,