A number of key features distinguish successful groups from less successful conventional organisations:
- Quality of mission. Is the mission clear, long-term, and impossible? The mission must be bold and clear enough to attract and hold the very best participants. Confused, shallow, over-complex, or inaccurate missions act as a barrier.
- Appropriate funding. Does the group have a benevolent patron or some backing from business? Without sustainable long-term funding, the project is a non-starter. The most successful groups have healthy relationships with profitable businesses.
- Freedom of access. Can anyone join the project, at little or no cost? It must be very easy for people to join and contribute, even anonymously. This guarantees diversity of opinion, an essential ingredient for a successful group.
- Unpaid membership. Are contributors volunteers, or professionals? A professional organisation team is good, but the bulk of the work must be done pro-bono to ensure independence of opinion. Most professionals - with some exceptions - care about their next paycheck more than their work.
- Non-exclusive membership. Are contributors loosely tied to the group? I.e. do they "belong" to the group? If they do, they become emotional and this works against logical analysis. The group must not become a primary badge of identification. I am not a Wikipedian, I simply contribute some articles.
- Well-written rules. Are the rules easy to understand and sensible? Clear rules resolve most disagreement before it happens, and allow people to trust each other implicitly. The first rules to be defined are the rules for making and changing the rules. The first authority is the authority to make and enforce the rules.
- Vandal-proof. Is the group equipped to detect and punish vandals and recover from damage? The most robust groups assume that everyone is biased until proven otherwise, and work accordingly.
- Tools, standards, processes. Does the group work on-line using good tools, standards, and processes? The quality of the tools is key to fast aggregation of information into knowledge. Tools are a key part of the investment in creating a group.
- Freedom to choose tasks. Do individuals work on what they please? If so, they will organise themselves optimally, faster than any manager can assign them to work. Identifying the problems, allocating resources, and monitoring success are all better done by the group itself than by top-down managers.
- Someone is in charge. Is there a clear and respected authority? There must be an authority who is ultimately responsible for the mission. Larger groups cannot work without strong authority. The best groups have benevolent dictators that never tell people "what" to do, but are absolutely strict about enforcing the rules on "how".
- Healthy growth. Is the group growing and attracting new members? Without growth, the community will die. Growth requires marketing, especially of competence: contributors are attracted to success.
- Decentralised workplaces. Is the group geographically spread out? Geographic separation and diversity is essential to wise decision making. Additionally, it is essential to have local teams that can focus on national issues.
- Open and transparent. Does the group work in the public eye, or behind the wall of secrecy? Digital society does its best work in the public eye, and secrecy inevitably causes incompetence. Privacy is necessary in situations of conflict.
- Free workspaces. Does the group offer free workspaces to anyone who wants them? This means, creating arbitrary wikis, lists, etc. If so, individuals will invest personally in the campaign. It must be extraordinarily easy to create and use workspaces.
- Internal maturity. Does the group have time to solve its own internal problems and challenges? All groups, whether they are five or five million individuals, take time to learn how to work together. Usually, it takes 2-3 years at a minimum.
- Strong identities. Does the group have a robust way of identifying its members? A group needs ways of identifying key individuals and punishing trouble makers. Anonymous contributions are welcome but are explicitly identified as such. People cannot change their identities.
- Local ownership. Do local teams and individuals fully own their work? Contributors must know they own their local projects and have full control over them. This gives people much more incentive to invest.
- A name on every brick. Are the smallest units of work properly attributed? Contributors must be credited for their work, this is, together with recognition and promotion, the main compensation for working pro-bono.
- Power structures. Are there proper tools to rank individual members? Plutocracies do not produce wise results, only meritocracies do. Contributors must be ranked, and promoted as they gain positive feedback from the community, and such rankings must be public.
- Metrics for success. Is there a clear competition for "karma" or "high scores"? Between individuals and between subgroups, competition for public prizes - the strongest type of recognition - is a key motivator.
- Content is easily remixable. Does the group have clear rules that encourage remixing of content? E.g. Creative Commons licensing on all content. The ideal remix license does not allow people to privatise content. It is important to formally promote sharing and discourage privatisation and parasitation.
- Neat structures. Does the group have a neat, predictable, documented structure? The structure needs to be predictable, so that we don't need to learn the whole map, just the general pattern. Breakdown by language and subject are typical.
- Sense of humour. Is the group able to laugh at itself? Humour defuses tribalism and emotion, and lets people work together even when they have huge differences. It is an essential tool for resolving internal conflicts.
These criteria are a subset of a list defined by Pieter Hintjens in 2007. The theory is that groups organize better when individuals have certain freedoms. You will probably recognize these criteria in many successful collaborative projects, from Wikipedia to the Linux ecosystem.