Month: July 2018

WIP.CHAT

Work in Progress is one of my favourite new communities. Members can share what they’re working on, get help from others, and be accountable to other members to keep their streaks going (were it not for the membership fee, they would have tens of thousands of members by now).

The community began as a tiny Telegram group with a few friends sharing what they had done or were working on. The accountability part took off and has now morphed into an entire community for people to share what they’re working on. They can earn streaks for posting several days in a row.

Work in Progress serves as an important lesson too about the range of possibilities for building a brand community today. It’s a place where you share your current projects, get help from others, and see projections in motion rather than the completed set.

Three lessons here are important.

First, test any concept in a small group first. If you can’t get the behaviors here, don’t build the community yet.

Second, if you want to enter a very competitive space, another forum-based platform doesn’t cut it. You’re going to need to build your own technology.

Third, communities that produce end results (actions) are better than members simply talking. Once you go beyond discussions as the main goal, you have an infinite number of possibilities for what the community can be.

There is no sector, even technology (by far the most popular sector for building communities) that’s too competitive to build a community. If you can identify the right use case, you can build a community that no-one has even tried to build before.

Research The People Whom Aren’t Active Members

July 30, 2018 ,,Comments Off on Research The People Whom Aren’t Active Members

A recent consultation in a client’s slowly dying community showed members wanted the community to remain exactly as it was.

This makes sense, if they’re participating they’re usually somewhat happy with the community. It’s hard to imagine something better. They might want a few small tweaks, but small tweaks won’t drive the community forward. Worse yet, it won’t change the trajectory of the community’s downward direction.

This is the problem with asking members what they want, they often describe what they already have (with a few minor tweaks)

The problem, of course, is you’re researching the wrong audience. It’s not your (happy) members you need data on, it’s the audience you want to expand into.

If you want to grow beyond your current audience, your research needs to go beyond your current audience. Find the people who have left the community, people in the sector who never joined, and people from related sectors who might be good members.

Uncover their problems, wishes, and concerns. Look specifically for rapidly rising trends. If 1% of members were using a new technology last year and 4% are using it this year, that’s a 400% increase. It might be worth paying attention to that.

If you want incremental improvements, ask your active members what they want.

If you want to increase the number of members who participate, ask inactive (or less active) members what they want.

If you want to expand, ask non-members what they want.

The Hidden Costs Of Pursuing High Engagement

July 29, 2018 ,,Comments Off on The Hidden Costs Of Pursuing High Engagement

I recently joined a new Facebook group.

The next day, I was welcomed with an @mention alongside 50+ other people. Every reply generated a notification which led to further replies. So far, I’ve received over 100+ notifications. It’s irritating, irrelevant to my problems, and exhausting.

From an engagement perspective, this is a big success. The stats are rocketing up. People are introducing themselves to the group. But, as with all efforts which chase engagement, the costs outweigh the benefits.

The immediate quantity of activity has driven me (and likely many others) from the very group I found interesting in the first place. I’ve already blocked the group from further notifications. From now on, if I want to know what’s happening, I’ll need to remember to visit the group. That’s not likely.

Since the hidden costs (the people who have been driven away by an extreme level of activity) never show up on stats it’s easy to assume it doesn’t exist. It’s a hidden problem.

A big myth of community development is members want to be in super-active hubs of activity. But anyone who has opened an inbox to 150+ emails, seen 300+ notifications in a WhatsApp group, or tried to follow any active Facebook group knows that’s not true.

I recently asked our FeverBee community what they would like to see going forward. The common thread was they didn’t want the site to be more active. They liked they could follow discussions, catch up on contributions, and participate. I doubt they’re alone.

Communities which chase the most activity typically become places filled with the people who have the most free time, the most passion or are most eager to build their reputations. These aren’t usually the best people. It looks good on the all the stats…except the stats which matter.

Members want to be able to follow and easily find the discussion that matters to them. They want the community to be relevant to them. Asking every member to introduce themselves can work when you’re small, but we already know @mention lists can do more harm than good.

A better approach would get them excited about the community. Highlight key members they might want to follow, share the best expertise ever created in the community, and make sure they know the community is a place they go to resolve an immediate problem they have or opportunity they need to pursue.

When you associate your community with quality you get more quality, if you associate it with quantity, you get more quantity. The future lies in the former, not the latter.

Beyond Information

July 26, 2018 ,,Comments Off on Beyond Information

Getting good, fast, answers to questions is the backbone of communities, but if you want people to stick around beyond that you need to offer a lot more than good information.

There are things only a community can offer: This includes:

  • Feeling a part of a unique and important group of peers.
  • Knowing your peers have your back.
  • Feeling respected by others.
  • Feeling like you make unique, useful, contributions.
  • Enjoying the experiences you share together.
  • Knowing your group is widely respected by others
  • Knowing your skills have progressed because of the group.

If you’re going to go down the information route. Do it. Really do it. Build the biggest and most incredible database for your topic ever. Do your keyword research. Find out what members really want. And completely dominate that sector. Start discussions about it, solicit contributions on those topics, build something new, different, and remarkable.

Brand Communities Have Inevitably Changed

July 24, 2018 ,,Comments Off on Brand Communities Have Inevitably Changed

Bit of a rant today.

As our list of brand online community examples shows, the nature of a brand community has completely changed in the past two decades.

Until recently, brand communities referred to devout superfans of companies like Apple, Jeep, and Harley Davidson gathering together to celebrate their founder, revealing a revolutionary new technology, help one another customize their vehicles, or participate in long journeys with one another. But brand communities today are completely different. They’re less about superfandom, creating warm, fuzzy feelings, and instead about showing a clear, visible, impact.

Making people feel great is nice, but being able to show millions of dollars in costs saved, new customers attracted, or new products launched as a result of the community is what matters. A happy, chatty, community which doesn’t deliver any obvious results is a luxury few companies can afford (and nor should they). It’s not a ridiculous question to ask “what’s the purpose of all this activity?” if the answer isn’t ridiculously obvious.

Fortunately, the potential of brand communities today is bigger than it’s ever been. Almost all brand communities have moved online and can reach far more people than ever before.

They create resources and host discussions read and watched by millions. A useful tip shared over the barbecue at ‘Camp Jeep’ might be heard by a dozen people. A tip posted in the Fitbit community can be read by tens of thousands of people.

Better yet, it can be tested, improved upon, and become common knowledge amongst all. It can drive forward innovation in ways which are scarcely imaginable.

Most members of today’s communities are less likely to be seeking a deep sense of belonging and more likely to be looking for something they need right now. This might be expertise, a place to build their reputation, or a chance to help others and feel good about themselves.

Companies big and small too often lack the ambition for what a community could be and the determination to see it through. As the potential of brand communities has grown, the ambition of their creators has shrunk. It’s a lot easier to launch another Twitter account than to take a group of customers aside and build something special.

Managing a community, and being responsible for a brand’s best and most valuable customers, should be one of the most exciting and important jobs in any business. It should be a job others aspire to and covet. But too often the task of engaging customers online falls to a junior staffer with limited experience and worse prospects.

Too often these staff are forced to chase meaningless measures of engagement rather than forging an indispensable community among their members and their colleagues.

It’s impossible to build an indispensable community when we’re forcing members to choose between clicking ‘like’, ‘share’, or ‘comment’.

Brand Communities Are Frequently Sabotaged By Bad Audience Research and Analysis

July 23, 2018 ,,Comments Off on Brand Communities Are Frequently Sabotaged By Bad Audience Research and Analysis

A prospective client recently asked:We know our audience better than you, we can do the research ourselves, why would we need your help?”

It’s a really good (and fair) question. We will never know an audience as well as you. You’ve been working with them for years.

But research isn’t about what you know, it’s about what you don’t know. Better yet, it’s about distinguishing what matters from what doesn’t.

The curse of knowledge is very real here and people tend to repeat the same mistakes. These mistakes usually include (in order):

  • Using research to support rather than refute existing ideas. By far the biggest problem is gathering data to support, rather than refute, existing ideas. Science is the process of coming up with a hypothesis and trying to disprove it. I’d estimate around 75% of people sabotage their own audience research trying to support whatever idea they had originally for the community.
  • Averaging the answers. This happens when you build a community around the average of responses instead of around a topic a small niche of members are intensely passionate about. This leads to [communities about topic] which struggle to gain traction instead of “Communities for data scientists who want to take data visualization to the next level”. You’re not looking for a community everyone finds acceptable, you’re looking for the thing that lights a spark under a small segment of your audience.
  • Homogenizing the community. This is the act of building a single member profile. This is silly. You have newcomers, experts, part-timers, hardcore advocates, different geographies, and personalities. The research phase should identify the different groups (we usually have 2 to 5) and the core passions, fears, and motivations of each.
  • Treating interviews like surveys. If you’re asking all members the same questions without diving deeper into responses, clarifying answers, finding out what they get really excited about and picking up on non-verbal cues, you’re basically running a very inefficient survey. Interviews are about getting to the deep emotional reasons to why people do what they do.
  • Bad data analysis. We correlate the time a community manager spends on each activity with the results they’ve achieved from it. Usually, we can cut 75% of activities without any noticeable impact on the community. They can then spend this time working on big wins. Most people collect a lot of data but miss the key correlations/analysis that turns it into actionable insights.
  • Not developing and testing options. The end result of this isn’t a big document, but validated steps to pursue. The analysis reveals several options to test, but the process of testing them takes time to get to the right answer. The end result is ideas which have been tested and are ready to go.
  • Collect data you don’t need. This happens often in surveys. Organizations try to drop questions in which are interesting but have no obvious influence on how we develop the community.
  • Breaking the law. Laws vary by country (especially GDPR in Europe). More than a few organizations have accidentally broken the law in the audience research phase.

You can certainly go through this process without outside help. We’re certainly not cheap and I doubt others doing this work are either. Our rates for this process typically range in the $20k to $35k region for a project that takes around 6 to 8 weeks.

However, this sort of research is the difference between a community which explodes to life and one which struggles to gain traction for years. It can save you hundreds of thousands of dollars, and many years, trying to get it right. It helps you ensure you’re building the best community possible. It’s a community which is almost guaranteed to succeed.

If you do go it alone, treat it internally like a $20k to $35k project. Smart companies know the success of the community doesn’t hinge upon the technology, it depends upon deeply understanding the needs of members. So give it the gravitas it deserves. This process takes time and requires more than just a lone community manager trying to get it done in her non-existent spare time.

Make sure the end result of the research is a powerful community concept which has been tested with the audience and shown to really resonate/drive participation. Don’t launch the community until you’ve validated a powerful concept and proved you can sustain activity on a small scale.

Dealing With Cynical Members

July 22, 2018 ,,Comments Off on Dealing With Cynical Members

A client’s community members are cynical, extremely cynical. They trash every new idea, criticize every decision, lower the tone of every discussion.

Cynicism is a hard stain to remove.

Cynicism is a self-preservation measure. You can’t be disappointed if you always expect the worse. If you can persuade others to agree with you, you feel validated in your negativity too. Cynicism happens when we’ve been let down, hurt, or feel angry about issues we often struggle to control. When this anger isn’t dealt with it festers and becomes cynicism.

You don’t tackle cynicism with excessive optimism. To a cynic, that’s not a breath of fresh air, it’s an irritating bug needing to be crushed.

You tackle cynicism by providing a place for people to identify and deal with the underlying anger and frustration. This place needs to be non-judgemental, emotions need to be reciprocated, and members need to feel genuinely listened to. Members need to bond closer together as a result of their shared frustration, hurt, and anger.

Once they’ve been listened to and responded to, you can begin to make small promises and keep them. Then you can make larger promises and keep them too. You gradually bond the community into a tighter group through their shared emotions.

The secret to tackling cynicism isn’t optimism, it’s a stronger sense of community. This is a community you can forge by providing members with a mutual sense of trust to speak openly about their anger, fear, and hurt.

“Let’s See How The Community Reacts”

July 19, 2018 ,,Comments Off on “Let’s See How The Community Reacts”

…is something you and your team should never have to say.

Our clients don’t see how the community reacts to anything. They already know. They’ve run the idea passed a decent-size group of members, they’ve refined it, they’ve gotten the buy-in from the top members (if they don’t, they don’t launch).

Usually, they’ve persuaded a small group of members (5 to 7) to share it and promote it positively within the community.

Dropping a new concept (platform, project, or anything else) might work well for record artists, but I wouldn’t recommend it for the rest of us. Most people don’t know whether something new is good or bad. They use the first reactions of other members to form their own opinions. Once that opinion has taken hold, it can spread rapidly.

So either don’t surprise members with something new (big changes should always be anticipated) or ensure you have respected community members engaged in the process and lined up to promote it (even better, do both).

It’s tempting to rush the launch of something new, but slow down a little. Get feedback, run the idea past people, and adjust your plans accordingly. Otherwise, the very project you’ve spent countless hours working on can hurt the very community you’re trying to help.

6 Models To Nurture Community Leaders

July 18, 2018 ,,Comments Off on 6 Models To Nurture Community Leaders

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.

1300+ Examples Of Successful Online Brand Communities (and you can add yours)

July 17, 2018 ,,Comments Off on 1300+ Examples Of Successful Online Brand Communities (and you can add yours)

Today we’re launching a database where you can explore 1300+ examples of successful brand communities and share your own efforts.

Getting good community examples has always been a challenge. Many people work in isolated silos with no ability to see what others like themselves are working on. For the past 18 months, we’ve been gathering data, cleaning data, and working to find a way to display the best brand communities on the web.

We’ve highlighted what we believe to be the top 10 communities by 4 different sectors and over a dozen of the most popular community platforms. In the coming months and years, we’re going to frequently update our ranking of communities and add to our database.

If you want to see your community featured, please add it to the database (you can also help us by suggesting new categories you would love to help create). You can also claim your own listing and keep it updated.

You can find our examples of brand communities here.

We hope you find it useful.

Upward cycles

The community sends remarkable stories and case studies to the PR team. The PR team promotes the stories which drives more people back to the community. This generates more remarkable stories to share with the PR team.

The community sends leads to the sales team. The sales team send prospects/new customers to the community. This creates a bigger community which generates more leads to the sales team.

The community sends innovative ideas to product engineers. Engineers start soliciting the opinions of community members more frequently. This drives better engagement in the community and more innovative ideas to engineers.

The community begins reducing the workload on customer support agents, who begin driving more people to the community, which further reduces the workload of customer support agents.

The top 5 questions each week are sent to the content marketing team. They create content based upon these questions with links back to the community for more discussions. This leads to more questions for the content marketing team to answer.

There are plenty of upward cycles you can start today.

Pay The Cost of Tough Compromises

July 12, 2018 ,,Comments Off on Pay The Cost of Tough Compromises

A potential client explained they would have to use a community platform neither of us liked because it was part of a CMS package they had just agreed to use.

Another noted their staff wouldn’t be allowed to participate in a community and they couldn’t let members talk about some of the most controversial product topics.

It’s incredibly tempting to agree to concessions like these to move a project forward. Often they’re presented as immovable facts which are impossible to change. But these are community-killing compromises. They will always come back to haunt you.

Organizations are far more flexible than they often admit. It’s down to you to do the hard work of building the relationships internally, putting together a clear (visual) case, and then standing your ground.

For sure, adapt the community’s goals, its concept, and many of the tactics to suit the organization’s objectives, its audience, and its resources. That’s the nature of collaboration. But never use a terrible platform, agree to topic-restrictions, or rules which are clearly going to hurt the community.

©2020 FeverBee Limited, 1314 New Providence Wharf, London, United Kingdom E14 9PJ FEVERBEE