Once the community hits the maturity phase of the community lifecycle, it needs people to help lead sub-groups. Without them, you get trapped with a small group of members driving all the activity.
To nurture leaders you need to resolve the tension between their need for competence, autonomy, and relatedness against your own need to have enough control to ensure they don’t tarnish your community.
The most common mistake is to enable everyone to create groups. This leads to a lot of dead groups littering the community. There are better models out there:
Model 1: ‘Anything Goes’
This is the Eve Online/CCP Games model. Anyone can create a corporation in the game, build out their own communication channels (Discord, Slack etc…), and run their fiefdoms as they please. The game’s creators, CCP Games, let natural selection work its magic.
The upside is leaders get full autonomy and dead groups don’t appear in any single list. The downside is CCP Games has very little control. Most discussions happen outside of the game. Leaders can (and sometimes have) tarnished the brand in a channel the game’s developers can’t control.
Model 2: Enable Anyone To Create Groups On Site
This is the most common model, best encapsulated by Reddit. Reddit provides a popular platform which attracts people to create subreddits (groups). Reddit exercises the fewest possible restrictions on the site. Anyone with a 50+ Reddit score can create a group. These subreddit leaders are largely left to build their subreddits as they please. Multiple subreddits can be created for the same topic.
This provides leaders with a large degree of autonomy but can lead to a lot of dead groups, protests when Reddit tries to enforce reasonable restrictions, and a confusing experience for visitors confused about the wide variety of site naming rules.
This works if you have a very popular platform to begin with, a culture where people do expect to join multiple groups, and people hear about the groups outside of the community (i.e. people don’t browse for the right groups to join).
Model 3: Back The Proven Winners
This is most recently the Facebook model. Anyone can (and does) create a group, but Facebook will invite the leaders of the best groups into private circles, provide training, and offer scholarships to the best of the best.
This supports the people who have proved they can succeed. Leaders have a huge degree of autonomy. But it might miss great leaders in less mainstream topics and doesn’t resolve the problem of dead groups. It does provide Facebook with considerable influence over their leaders.
Most importantly, it concentrates support in the areas where it is likely to have the biggest possible impact.
Model 4: Pick Great Potential Leaders
This is the Wikimedia Foundation model. Anyone can come up with a project and apply to Wikimedia for support. If the project matches Wikimedia’s goals, it gets financial backing to cover expenses (e.g event venues for edit-a-thons).
The upside is this provides WMF with influence and direction over their leaders. Leaders can thrive without them, but WMF grants guide contributions to the areas with the biggest impact. Leaders also retain a great degree of autonomy (they build their own sub-groups off-site). The downside is it can waste resources if leaders fail.
This is the easiest model to follow. Any brand can give support (promotion, technical, financial, expertise) to people who apply for it. But you need a very popular community to make it work.
Model 5: Create a Limited Number of Roles
This is the Nextdoor model. Nextdoor restricts the number of groups which can be created but allows anyone to create one. However, you must prove you can attract people to the site, drive activity, and keep things going or you might be replaced by someone else.
This works because Nextdoor essentially leases leaders land upon which they can build a community. However, because land is both scarce and popular, Nextdoor can enforce fairly rigid rules on the design of the groups, force leaders to prove their worth, and replace those which aren’t doing well.
This approach trades a leader’s autonomy (they have to abide by fairly tight constraints on the platform) with an opportunity to have an impact. The downside is it requires a structure with a finite number of groups and only 1 group can be created per topic. This in turn requires a reasonably popular community to begin with.
Model 6: Find The Best Leaders (filter out the weak)
This is the famous (or infamous) StackExchange model where potential creators can spend months, even years, making their way through the geekily named Area51 gauntlet to create a new site.
The process is complicated, but essentially members with good scores need to propose a site idea, define what it will be about, gain committed members for it, initiate and sustain discussions, and gather momentum before it gets added to the community. One site has been in Area51 for 8 years.
This approach flips the traditional model on its head. Instead of trying to attract leaders, StackExchange tries to drive them away until only the best are left standing. The advantage is it results in only well-run communities which add value appearing on the site. The downside is the brutal gauntlet might drive away many who could have been nurtured into great leaders.
Which Model To Pick
The best model depends on your site’s popularity, goals, and the kind of members you have. This table below might help:
| Who can lead? | Level of support | No. of groups | Pros | Cons | |
| Anything goes (‘Eve Online’) | Anyone | Tech only | Unlimited | Leaders have full control, low cost | No control over leaders who can harm the brand by association |
| Enable everyone (Reddit) | Members with a 50+ Reddit score | Tech/social | Unlimited | Supports people who want to maintain a good relationship | Doesn’t focus on those who can have biggest impact High cost |
| Support the best (Facebook) | Anyone | Best groups get promotion and expertise | Unlimited | Supports people who prove they can succeed Medium cost | Great leaders might slip through the cracks |
| Pick the winners (Wikimedia Foundation) | Anyone | Financial | Unlimited | Biggest impact for resources Medium cost | Can waste resources if they fail |
| Create limited roles (Mozilla) | Need to apply | Training program/full access | 200 | Retains power Medium cost | Limits potential no. leaders |
| Iron grip (StackExchange) | Prove they can lead a group | Promotion on very popular site | One-group per topic | Retails power and keeps only the best leaders | Deters many great leaders Very high cost |
The bigger your audience, the more you can set strict rules around participating.
Enabling everyone to create a group is probably a bad idea. But if forced, you need to decide whether to support the good or filter out the bad. Try experimenting and see what works for you.
