It’s a lot easier to launch a new community if you have a lot of customers or thousands of names on your mailing list.

If you don’t, you’ve got a few approaches you can take:

Option 1: Grow an Existing Audience, Then Launch

Most common for founder-created communities is to build a large number of existing contacts in the 3 to 6 months before launching the community. Get them to engage in events, through other groups, and steadily grow from there. Most of the communities you see followed this approach (albeit unintentionally).

Option 2: Pay for Advertising

If you don’t have an existing audience, you can pay for advertising.

Social ads can work well here. Each new member will cost about $5 – $10. Invest a few thousand dollars and you will get 300 to 600 registered members.

Assume you can keep 10% of them engaged, and you have your 30 to 60 founding members to get started. Keep investing a few thousand a month and you will start to see steady growth.

It’s not cheap for many, but it can work out cheaper than spending months building your own audience (your time costs money too). It can also be a lot quicker.

Option 3: Start A Tiny Group and Grow Steadily

Create a group on Facebook, Twitter (hashtag) or another channel that people already use and use built-in network effects to attract members. You can test different concepts and ideas, then grow steadily. Only launch your community on these groups. Once you have some engagement you want to move across to something you control. Remember at any time these platforms can seize control of your platform, reduce your reach, or remove it entirely without warning.

Option 4: Create Something With Viral Power

My favourite (and least used) option. Create a community so remarkable and different that people can’t help but talk about it. Do what Kaggle did for data scientists and Figure1 for doctors. A simple forum probably won’t cut it. Target 1% of your prospective audience who are ridiculously excited or challenged by an aspect of the topic. Then overwhelmingly cater towards it.

Feel free to have crazy rules that no-one else would dare to use. Perhaps ban unverified opinions, force members to mentor at least one other, or perhaps ban questions entirely and force people to share what they’re up to in some other medium. There are endless possibilities.

Select whichever approach makes the best sense to you.

Of course, if you want to grow really fast…select all four.

I have a list of organizations who turned down our consultancy proposal for being ‘too expensive’.

…and then sunk an additional $250k to $1m+ into their community (staff + platform costs) without managing to make it work.

That’s a lot of time and money to waste.

Many of these communities today have no organic growth at all – the community manager just pushes out new content and discussions hoping something will miraculously happen and the community will spring to life.

The miracle is never going to come, it’s just more time and money down the drain.

Whether from us or from someone else, get some help. It’s tremendously painful to watch brands repeat the same, easily avoidable, mistakes and condemn potentially incredible communities in the process.

If your community isn’t as successful as you want it to be, get another perspective. Get someone to challenge you, push you, and give you a wide-angle lens detailing what other organizations like yours have done to succeed.

Get someone who can explain other ways to figure out what your members want, build support etc…Get someone who can explain how to restructure your community, better design it, and ensure it’s making the best use of your staff/technology investments.

No, this might not come ‘cheap’, but when you’re spending $250k+ on your community a year you probably don’t want cheap – you want results. If getting consultancy support turns the community around (as we’ve done, consistently), it’s a bargain.

The downcycle is pretty clear.

It’s a classic engagement trap. To get more engagement, you dumb your content down and make it easier to participate. You make it sillier, more fun, more emotive, more controversial, more clickbaity, and wait for the clicks to roll in. Instead of asking for thoughtful comments you aim for likes, clicks, and immediate reactions. To keep engagement rising you need to dive deeper and deeper into the engagement trap.

The opposite is an upward cycle. You set high standards and consistently raise them. Towards Data Science is an impressive example. To be published, you need to undertake your own study (hours and hours of work) and submit your findings. You can either tackle new problems or tackle existing problems in a new way.

When you set high standards and enforce them, a successful submission becomes a badge of honour. Others want to be published too. As more contributions are submitted you gradually raise the bar (keep the total number of contributions you publish restrained).

Towards Data Science isn’t alone, ProjectManagement.com and others have proved if you want to build a really powerful community, don’t lower your standards – raise them.

It’s a lot harder to start an upward cycle than a downward cycle – but that’s what makes it a lot more valuable. When your competitors start lowering their standards, start raising yours and stand firm.

…you might want to consider the risk factors.

When you enable members to create groups on your community you’re essentially renting them your brand name (and a small portion of the audience) to pursue their own goals.

A few things to consider:

1) Which members will you allow to create groups? Can anyone do it or only a select few who have demonstrated the ability to manage a group (hint, choose the latter)?

2) Do you have a training program for members to create groups?

3) What happens if group leaders don’t crack down on minor abuse?

4) What happens if group leaders don’t crack down on serious issue?

5) Will you allow multiple groups on the same topic or only one group per topic?

6) What happens if a group leader becomes inactive (who can replace her and what does this process look like?)

7) What happens if a group leader is active but the group isn’t? What level of activity does a group need to continue as a group?

8) How will you remove groups which don’t take off without upsetting group leaders or members (and what will you do with the content in that group?)

9) Who gets to name the group? What are the restrictions on group names?

10) What happens if groups are local and want to meet in person? What is your legal liability for what happens? To what extent can you support groups?

11) What happens if group leaders get together to demand changes you are unable or unwilling to make?

When groups work well, everyone wins. When it doesn’t, you can upset your best members, building hundreds of ghost-villages, and see your brand name tarnished.

If you don’t love research, identifying costs, project planning, developing benchmarks, getting internal support, and building decision trees, don’t become a community strategist.

We’ve created more community strategies (and trained more strategists) than most.

Believe me, this isn’t just a small jump to focus ‘on the big picture’, it’s a completely different type of work.

It’s often work community managers discover they don’t enjoy.

In our coaching, four areas seem to surprise people the most.

1) You need to cost your strategies.
If you’re presenting a plan and you don’t know the resources it requires, you have no idea if you’re creating something feasible or not. Worse yet, you’re not taking your work (or your colleagues) seriously. It’s hard for people to support something if they don’t know what the costs are.

For example below, every community strategy we create is fully costed by both the time and financial resources required:

This is critical for two reasons. First, once you know the time required, you determine how many staff you need to reach each new level. This, in turn, guides you on the financial resources you need. Second, this makes the plan flexible. You can present options based upon resources available. (i.e. ‘with [x] resources I can achieve [y], but with [xx], I can achieve [yy]’).

Once you have your tactics prioritised you can quickly adapt them by the time and resources you have available.

2) You either have research or guesswork.
If you can’t point to the research that supports every assumption in your strategy, you should be honest and call it guesswork. You have no idea if it will work or not, (but, hey, at least it sounds good).

You have to love the research side of community. You have to enjoy interviewing dozens of people internally and externally, identifying segments, understanding their priorities, and using that to craft your approach from the community’s goals down to the specific tactics you decide to use. You have to analyze in-depth what is and isn’t working (more on that below). You have to enjoy reaching out to peers and researching other communities to identify the best approach towards everything you want to do.

3) Building Decisions Trees > Reporting Metrics
It’s one thing to set measurable goals and KPIs on the way towards achieving them. Reporting what happened/building a narrative around the data is vital. But reporting doesn’t tell you what you’re going to do differently. If you want your strategy to be more than just a snapshot in time, then you need to build decision trees based upon what the data tells you.

If a metric you really care about drops by 10% in the next few months, what will you do differently? You need to build out what success/failure of each tactic looks like, what you will stop doing and what you will invest more resources in doing based upon the data you’re seeing.

This is how you build a strategy that lives indefinitely instead of a strategy which becomes stale from the date it’s published.

4) Building Internal Alliances
You should never ‘drop’ the strategy on surprised colleagues.

The strategies we present are never a surprise. They’re the summation of a lot of conversations and collaborative decisions we’ve guided clients through to reach a point of agreement. If people are disagreeing with aspects of your strategy, you’ve probably not communicated frequently enough with your colleagues.

Anyone can whip up a detailed strategy document in a dark room in a week or two. The reason it takes us 3 to 4 months to build a strategy is we bring our clients along the journey with us. Every point is discussed, objections highlighted, concerns addressed as early as possible.

The end result is a strategy which has the support of all the key stakeholders (and key members of the community). There is a gulf in difference between tepid acceptance and enthusiastic support of a strategy.

Once the strategy is established, you have to continue to maintain strong relationships with stakeholders, address concerns, demonstrate results, build a shared narrative through powerful stories etc…

If this is the kind of work you want to do, then, by all means, push to reach the strategist level. But be aware it’s a very different kind of work from managing a community.

…and if you really want to thrive, don’t wait to reach the strategist level before acquiring these skills. Gain these skills before you have the job.

The best answer to a member’s problem might not be in response to a question posted in your community. It might be in documentation your company has created already but never duplicated in the community. It might be published in the help centre. It might even be shared in another social network/community (reddit/StackOverflow/GitHub/YouTube etc..)

If your search bar only retrieves information from your community, you’re limiting the ability of your members to find the information they need.

The community managers at the biggest communities have long realized that native search (the search function that comes with the platform by default) doesn’t quite cut it. Native search doesn’t typically let you:

• Query multiple databases and retrieve the answers which best match the query.
• Let you boost the best/most updated answers above others or older content.
• Create keyword synonyms showing relevant results even if members don’t quite know the answer.
• Identify and close content gaps (queries which don’t retrieve a satisfactory result).
• Use AI/machine learning to display the answers which best solve a member question.
• Show relevant content/queries alongside existing content/questions.
• Track call deflection using queries which prevent members opening a ticket.
• Analyze what members need in depth, and feed that information back to product, support and marketing teams.

Upgrading your search bar doesn’t come cheap (a license with Coveo/SearchUnify/others costs $20k to $50k per year), but it could also be a bargain.

It’s a bargain if it helps thousands of extra members a year find the answer to their question.

It’s a bargain if it also shows community solutions alongside others in the help centre.

It’s a bargain if it encourages hundreds of members to collaborate together to close the content gaps the tool has identified.

It’s a bargain if it accelerates how you develop products and gives you insights into exactly what your community needs.

It’s a bargain if it helps you to prove the incredible value of your community.

Your native search is fine when you’re just getting started, but if you’ve got a lot of documentation sitting outside of the community, have 50k+ members, and want to provide members with the best search experience, you might need to upgrade.

It probably doesn’t feel like it, but attracting bad actors is a sign of success.

It means you have a community that’s worth spamming, trolling, and hacking. It means you have a community where members care about what their peers in the community think of them and battle with each other to maintain their reputation.

The only real failure is a failure to plan for this.

Every large community has had to deal with the same problems you’re facing now.

Before you reach this size, you should have prepared scalable systems for dealing with spam, trolls, and hacking attempts. You should have learned from the many peers who have come before you and dealt with these issues.

You should have decided where you lie on the freedom of speech vs. protecting members from abuse continuum and be aware of the trade-offs you’re making.

And you should have a team that’s trained to resolve disputes effectively and enforce rules intelligently.

So, yes, congrats, you’re worth trolling…now don’t mess it up.

If you want more participation, an obvious area to target is to reduce the costs (that crippling anxiety) of asking a question.

There are many reasons why members don’t ask a question, they include:

• They’re afraid of looking inferior to others.
• They’re too invested as being seen as the experts themselves.
• They don’t think they will get an answer.
• They don’t think they will get a good/the right answer.
• They’re afraid of spamming the community with a question.
• They don’t know the right words to describe their problem.
• They don’t know if the question has been asked many times before.
• They don’t want to be in debt to others.
• They don’t want to exploit friendships.

The more effective solutions remove the problem entirely. You need the technology to let members:

1) Ask questions anonymously. Let members ask questions anonymously. If your platform doesn’t offer this, request it.

2) Let members set a time period for questions. Allow members to ask a question which only shows for a limited amount of time (or until they’ve accepted the answer).

3) Let members decide who sees the questions. Categories help, but it would be better to let members decide whether the question can be seen by the entire community, only those who have posted in a category before, or just a handful of top experts/close connections. Group @mentions can help here. Create groups of experts and let members tag in relevant groups.

However, if technology solutions aren’t an option, you need to eliminate the fears members have through persuasion. That might mean:

4) Tell members ‘top professionals aren’t scared to ask questions’. Help members to rethink asking a question from an act of weakness to an act of courage. Support boldness and the bravery it takes to ask questions – especially on really beginner topics. Reward members by the number of collective visits their questions receive.

5) Develop on-page nudges of great questions. See in the social design webinar how platforms like StackOverflow, Apple’s community and others provide nudges to help members ask great questions. Guide people through asking a great question that’s going to get great responses.

6) Tag newcomers to ask their first question. Most won’t respond, but we’ve seen you can raise the number of newcomers who ask a question by double-digit percentage points if you tag them in to ask questions they might be struggling with.

Getting more members to ask questions is one of the most intriguing challenges you can tackle. You need to get into the minds of your members, isolate their social fears, and then design effective solutions to tackle them.

An accepted solution is a solution to the original poster’s question.

A best answer is someone’s subjective opinion of the best response from a range of answers.

The distinction is important. The former is based on facts, the latter is based upon opinions.

The original poster is best qualified to mark an answer as an accepted solution. She knows better than anyone whether it solved her problem or not. No promise is being made other than it solved the problem for 1 person. If it helps others too, great.

The original poster is less qualified to select the best answer from several responses. That requires expertise in the topic (expertise they might not have if they’re asking the question). The premise is different. ‘This might not solve your problem, but it’s the best answer we’ve got here.’

Both can be abused and misused.

Two useful principles here:

1) Accepted solutions are ideal for support communities. But they need automated prompts to the poster to mark an answer as an accepted solution or allow other members to highlight if an answer worked for them (with a limit i.e. 3 tags) for that answer to be marked as an accepted solution. Customer support staff can skim other responses and mark successful answers.

2) Best answers are ideal for customer success communities with a focus on helping others learn from each other. Here it helps to allow top members in the insider/mvp/superuser program mark an answer as best.

Don’t fall into the trap of confusing what’s based upon facts and what’s based upon opinions.

Adding more is a good thing…until it isn’t.

Over time, it becomes harder for members to find the good stuff, follow conversations, recognise other members and know which events are worth attending.

A good rule of thumb is to spend half as much time pruning and cleaning what’s in the community as adding more to it.

An easy way to improve most communities is to spend some time taking down less relevant content, archiving old discussions, pruning inactive members, and removing lesser used features. Look at your event calendar too and cease events which aren’t clearly driving great results.

I’d remove:

• The bottom 10 to 20% of content (by landing page visits from search results).
• All discussions which have seen <5 visits within the past 30 days.
• Events which don’t attract at least 50 to 100 participants.
• Members who haven’t visited within the past year.

Ironically, removing stuff is as likely to help you achieve your goals as adding more stuff. It increases your search rankings, improves the member experience, retains confused newcomers etc…etc…

Subtracting stuff from your community can be as valuable as adding to it.

James reminded me this week they’re not the same thing.

If you really love a brand and what it represents, you’re probably part of a brand community. Harley Davidson, JetBlue, and Supreme all have strong brand communities.

Brand communities are forged through numerous hard choices at the highest level of the organizations over many, many, years. They have uncompromising values which their followers strongly identify with.

Very few of you are managing a brand community right now (and fewer have the power alone to build a brand community).

However, if you just want to fix problems with your phone, learn to use a piece of software better, or connect with others in the same situation as you, you’re probably going to join and participate in a customer community.

Customer communities begin with specific needs customers need to solve. Members will engage with each other to satisfy those needs (and those needs may evolve over time), but they’re not going to feel the brand represents their identity.

It’s easy to get confused between the two (and they clearly overlap in many areas). You can waste your time and your audience’s attention on things you think your members want to hear because they love their brand. It’s usually far better to focus that time on the immediate.

What are the problems your members face this week, how you can solve them better than last week?

What are the new opportunities your members can seize this week? How can you help them seize those opportunities?

It’s a different kind of work. No less valuable, but far more specific in the challenges of the day.

p.s. Podcast interview with Bevy.

Does it have any chance of doubling any metric that matters to you? (growth, activity, call deflection etc…?)

Does it strategically reinforce whatever it is that makes your audience unique? And thus is something any other community would struggle to replicate?

Have members expressed an overwhelming (unprompted) demand for it?

Will failure to do it cause the death of your community?

Have you decided what you will stop doing to invest the resources to make the new feature a success?

You better have a compelling reason to distract yourself from making your current community features better, divide your time and resources into smaller chunks, and expend precious audience attention on something that might not work.

(p.s. because someone else is doing it is both the most common and very worst reason to do it).