Category Archives: Tech

How you can avoid a community revolt like the one Reddit just experienced

These days I hesitate to dive in and criticize companies for their community-building missteps. Building community is hard, and we all fail at some point or another.

However, I can’t resist being about the 3023rd person to weigh in on the
Reddit protest
that occurred this weekend. Much of the analysis so far, while fine, has suggested a single action (the firing of Victoria Taylor) caused the event. Instead, it seems clear to me that this was simply the straw that broke the camel’s back. There are much deeper problems over at Reddit, and I think it’s worth analyzing them because we could all learn from their mistakes. Here are my thoughts on what went wrong and how they could have avoided a blowup.

A single point-of-contact for community is a dumb move. (It’s also a common one.)

Having a single point of contact for a community means that you are one step from failure. If that person leaves, or is let go, or acts like a jerk, or does any number of things, your community is immediately in danger.

If you’re at a company with multiple community employees, make sure they all engage with the community on some level. Certainly some will have much more interaction than others, but being a recognized face is important.

Unfortunately, at many companies there’s only one community hire, so this is impossible. In this situation, try to have a non-community person (maybe a marketer or product manager, or even an office manager…someone with people skills) at least occasionally interact with your community. Build a sustainable community integrated with your company. Which leads to…

You need to be part of your community.

By all accounts—including her own—CEO Ellen Pao spends very little time talking to her community on Reddit. The employee they eventually asked to step into the moderator liaison role does not seem to be very active either, and often seems to get into tiffs with users regarding moderation actions. Meanwhile, the celebrated employee that’s leaving seemed incredibly active (she even had a whole subreddit for her two cats). Being part of your community engenders trust and helps you better understand them. I suspect the backlash would be far less severe if they had been a bigger part of their community.

If you let go an employee who is crucial to any major community projects, immediately fill that void.

Part of the major frustration from the Reddit community is the fact that Victoria was the coordinator for Ask Me Anything (AMA), the Reddit Q&A series (and often was the point person for verifying celebrity identities). With her departure, many of these sessions across multiple communities were suddenly in jeopardy. Reports vary, but it sounds like Reddit didn’t step up to temporarily fill this position and ensure these AMAs proceeded successfully. In fact, it sounds like some AMA moderators learned about Victoria’s departure through AMA subjects who couldn’t get ahold of her. This is just lazy on Reddit’s part; they should have immediately had an existing employee step in to facilitate this important part of the Reddit experience.

Monetizing a community is tough. Involve them, help them understand why it’s important, and know that there is a line past which you will alienate them.

I don’t know the magic trick to monetizing a community. It’s hard to make money off of something that’s explicitly designed around quality personal interactions rather than transaction. Steps towards monetization can cause community members to get very worried and upset…sometimes for legitimate reasons, sometimes imagined ones.

With that in mind, you must involve them. You must help them understand. Many times, the issue is simply your community picturing a huge company with insane profits and swimming pools full of gold trying to squeeze a few more bucks out of them…when the reality is that the company is rag-tag group of passionate people struggling to become profitable so they can keep the site afloat. Communicating this and being transparent about why and how you’re looking at monetization will greatly decrease the amount of frustration the community feels.

Involving community beyond that can be tricky; if you directly ask the group if you should do x or y, they’re likely to say “no” (and then you’re in a tough place if you do it anyway). But two options are a) doing user testing (the real kind, in a room rather than deployed live to a bunch of unsuspecting community members) and b) organizing a community advisory board. Both not only help the community feel that you took their concerns into account, they also should help you make a better decisions.

Finally, you need to know that there’s a line. There’s always a line. The communication and collaboration you do with the community can push that line further, but there’s a point (whole page take-over ads, popups, advertisers who clash with community values, etc) at which the community will say “no, that’s not ok”. Traverse further at your own risk.

Ad revenue is built on large numbers of visitors. When visitors are a direct result of the work a minority of volunteers provide, that vocal minority is, and should be treated as, very important.

At a SaaS company you’re likely to try to ignore the vocal minority asking for power-user features that won’t help you sell any more licenses to average users. But with a community, that vocal minority is often the reason the community is functioning and healthy, especially when they have significant responsibilities and projects. It seems that Reddit has been taking their moderators for granted, doing a bad job communicating with them and building much-needed features.

As Drew Olanoff put it: “Moral of the story? Making bucks off of the backs of a community is possible, until they take their backs somewhere else. It’s a tightrope. Reddit fell off.” Reddit should be prioritizing moderators above almost all else aside from revenue, because without them the site will collapse. The fact that Reddit apparently promised new tools and features for moderators and then never delivered is proof that their priorities are in the wrong place. They should have dedicated developers working on tools for moderators and dedicated staff communicating regularly with moderators (and, post-crisis, they now seem to).

Every company should have a crisis plan that involves communicating immediately (even if your communication is “we can’t talk about this yet”).

Crises happen. You can’t help that. But you can have a plan. And the first part of that plan needs to be communication. When community members hear about drama from other community members, it often gets exaggerated. This is true in most contexts. When a coworker tells you that Jenny from marketing was let go, it usually includes gossip or drives you to speculate, often leading to worst-case scenarios (“I bet they’re shutting down the whole department!”). But if the company communicates quickly and honestly, much of this will be squelched.

Reddit did a terrible job of this, with very little public communication about the issue. Worse, Pao went on to talk to the press about the incident before talking to the community. (She did post a single, unimpressive comment in one of the public threads. She later complained that she was being “downvoted” so her message could not get through…despite obviously having access to the company blog and announcements subreddit.)

Letting go of an employee that’s extremely connected to your community should be done very carefully. Ideally, you should work with that employee to ensure a smooth transition. Most employees would prefer this; it reflects badly on them, too, if a community collapses after they leave, proving they didn’t build a sustainable community. Sometimes this isn’t possible. We still don’t know what happened with Victoria. Sometimes departures are not amicable, or there are legal reasons you can’t discuss them. But just like I said in one of my favorite posts for UserVoice, saying nothing is far worse than saying no. So if you can’t actually talk about the situation, just acknowledge that it exists and is important. “Hey all—we hear that you’re upset about this and hope to discuss it with you soon, but for legal reasons we can’t just yet. Stay tuned, and thanks for your concern and passion.” That simple move would have alleviated a lot of the anger.


Community is hard. People are fickle (especially on Reddit). Balancing mission and revenue is tricky. There’s no silver bullet. But there’s also no excuse for neglecting your community and then on top of that not executing a good crisis and transition plan. I hope we can all take a moment this week to review the above and make sure our own communities are not set up for failure like Reddit was.


Footnote #1: What should Reddit do next? I think Sam Houston has some great thoughts.

Footnote #2: Before anyone brings it up for me—yes, I absolutely think the misogyny that is rampant on Reddit had something to do with this. Their disdain with CEO Ellen Pao seems half rooted in legitimate frustrations, half in their disgust over her lawsuit. That is unpleasant and inexcusable, but I still think it was a minor part of the equation. The points I’ve made above are far more relevant.

Footnote #3: Thanks to Jennifer Sable Lopez, the Community Building
Stack Exchange
, various articles, and everyone on the CMX community for helping me polish my opinions here.

Sharing economy lessons from Lyft, Yerdle, & more

I recently had the pleasure of attending the San Francisco Community Manager Meetup‘s panel on the sharing economy and writing the official summary post. Here’s an excerpt:

“If there are 19 other drivers posting on the forum, you think ‘Ok, I’m gonna keep driving’.” Super-users were key to helping Lyft scale. Several months into a year in which their goal was launching 70 cities, they had launched exactly one city. They were struggling to find an office space and hiring local employees to work with drivers. Forced to innovate, they thought “what if, instead of having new drivers come into the office, we just have them meet an experienced driver in a parking lot?”

“It was scary,” David admits. “But it worked.” 

Check out the full article at CMX.

What if you subtracted “women” from that question?

For most* questions, if you replace “women” with “someone” and you get a different answer, there’s something wrong. Either there’s a institutional issue you should address, or you’re being sexist.

Take the case of Microsoft CEO Satya Nadella, who was asked what advice he’d give to women who want a raise (and might be uncomfortable asking). If Nadella was not sexist, he would have two options:

1) Give the same advice he gives to men: meet with your boss, show all the good work you’ve done, and ask.  This advice probably wouldn’t be super-helpful, seeing as women make 78% as much as men, but it would have been ignorant instead of sexist.

2) Acknowledge that there are a lot institutional problems with gender and pay. Talk about how women are often called manipulative or bitchy when they are strong-willed in the workplace, but similar men are considered driven and ambitious. Talk about how that needs to change, and maybe outline some ways that someone powerful like, I don’t know, the CEO of Microsoft, could change things. This also wouldn’t be a super-helpful answer, but it would be honest and definitely not sexist.

By telling women “It’s not really about asking for the raise, but knowing and having faith that the system will actually give you the right raises as you go along”, Nadella is being sexist.

You might argue that he doesn’t know about the huge wage gap between men and women. Maybe that’s the case (disappointing if so). But again:  if he thinks women get the same opportunities as men, why would he give them a different answer than he would give men?

As a straight white male, anyone I’ve ever asked for advice on getting a raise has told me to ask. You can find plenty of this advice on the web. You’re generally supposed to gather examples of your great work, do some basic research, and then ask. “It never hurts to ask”, “you’ll never get what you don’t ask for”, etc – this is the stuff I’ve heard many times.

Why didn’t Nadella give this advice? Because he treats women differently. And that’s sexist. Period, full stop.

Let’s hope this is a wake-up call for him. I know it’s a wake-up call to me and many consumers that to not buy any Microsoft products. In the meantime, I’d recommend women avoid applying  to work at Microsoft, unless you want to cross your fingers for your “good karma” to somehow net you a raise.


*I will admit that “My water just broke, what should I do?” should have a different response when coming from a man.

Ryanair is failing, and I couldn’t be happier

 Ryanair CEOThe problem with the balance between marketing and caring for customers is that marketing is often immediate. You can put an ad out and immediately see people visiting and signing up and spending money. You put effort into customer service, community building, or just basic user experience and you don’t see an immediate result. But us community professionals swear that you will see a long-term benefit.

That’s why I’m so pleased to note that Ryanair is seeing its worst annual revenues in five years. I’ve been saying for years that Ryanair’s tactics were going to explode in their face, but people kept saying “they’re selling tickets, so apparently is not an issue”. To me, this is fantastic validation that caring for your customers does make a difference and treating them badly does affect the bottom line.

But really, what would you do if there was no email? Thomas Knoll’s drastic experiment

We’re building the next generation of companies, and they look nothing like the previous ones. Right? Email is dead? Meetings are stupid? Companies should be flat?

I hear a lot of people talking about these subjects and some trying to implement them, but nobody as drastically and definitively as Thomas Knoll. His writeup on the email-less culture at Primeloop is both terrifying and exciting. Definitely worth a read!

“Community is a discipline”

“When community is your discipline, the core of your work is focused on understanding and putting into practice the development of communities. You might also understand and apply marketing, support or other practices but they’re traits, not your core discipline.”

David Spinks, as he often does, describes succinctly what I rant about regularly: Community is a discipline, and the more you try to make it a subset of other departments the more it fails. Other departments can adopt community-focused strategies (please!) but it will never make sense to say that the Community team is part of the Marketing team.

Interviews with me

A few very pleasant folks have interviewed me in the last couple of months, so I figured I’d share those here.

Support-Driven Podcast – Scott Tran interviews me about kicking off a community effort, finding your community niche, and combining customer service & community. Also available on iTunes.

Big Door – I discuss why customer loyalty isn’t marketing, the community management trends I see, and metrics one should look at.

Startup Product Summit – A little bit older, but arguably one of my best and most widely-accessible talks. I cover customer feedback, customer-focused product design, and how to understand when data or customers are misleading you. Slides available here.

Hope you find these helpful, useful, or at least entertaining! 🙂

How do you measure the erosion of your brand?

“This one promotion seems successful — but they have no idea how much money they are losing because millions of customers stopped opening their direct mail. How many people don’t trust the brand because they’ve been tricked (lied to?) about ‘important information’ five or six times a year? How many bank accounts not opened, mortgages not applied for, car loans taken elsewhere, business accounts moved away?”

From Are your profitable programs destroying your brand?

I think most community managers have struggled with this. I’ve dealt with this at every job I’ve ever had. If the benefit can be measured and the damage cannot, how do you make a solid case for ceasing the activity?

Any ideas?

The hidden cost of overwork

man sleeping at desk

Frequently I hear startup CEOs & managers boast (and even front-line employees complainbrag) about how they or their staff work endless hours. I’ve done it myself in the past. But building a culture of burning the midnight oil is bad for your company, for two major reasons.

#1: Burnout & Replacement Cost.

If your employee burns out and has to be replaced, it costs money (some say 20% of that person’s annual salary)…not to mention it makes them unlikely to recommend your place of employment to others.

#2: False Positives.

The more hidden and insidious cost is false positives.

amp to 11You have X amount of work each day and your Y employees are able to complete it within 24 hours. Success, a sustainable business! If it turns out, however, that in order to accomplish this baseline work people are consistently working very late hours, you have a false positive.

If your staff were really only using ~9 hours a day to complete this task, it would mean they have flexibility in case of a increase in workload. Because they are actually working their maximum number of hours, it means that any increase will be disastrous. Your employees wouldn’t be able to accommodate this new work, even temporarily, because they’re already past capacity (and there would likely be lots of burnout).

An increase in workload could come from any direction…something bad like a major bug or an issue with one of your partners, something good like the New York Times covering your company and creating a rush of customers, or something random & unexpected like an employee pregnancy or a hurricane knocking out your servers. One change and boom – you’re unable to meet demand.

If you create a culture of overworking you’ll only succeed until you are blindsided by something…at which point you will suffer. Promoting a culture of hard work balanced with realistic hours may cost you more up front because you’ll have to hire more, but it’ll save you a lot of employee turnover, bad reputation, and disastrous situations.

(Obviously – if you’re an early-age startup, this is probably not applicable. If you are a founder or CEO, sadly, this may not be applicable. Insert other necessary caveats here.)


Sleeping photo courtesy of Svein Halvor Halvorsen.
Amp photo courtesy of This is Spinal Tap. If you haven’t seen this, then we can’t be friends. Go watch it now.

The Secret Structure of Great Talks

Like most good advice, Nancy Duarte’s TEDx presentation on how to give a great talk is both obvious and enlightening.

In short: talk about the status quo, then the possibility of the future. Repeat as much as possible. End with the new status quo that you’re proposing.

The repetition is very key, and her example of the classic Martin Luther King Jr speech is especially relevant: pastors know how to use repetition!

I know that I too often I build presentations with the three-act model; I state the current situation, what I’d like to change, and what the results will be. I’ll be keeping this in mind next time.

I also  love her point about making the audience the hero. It’s easy to make yourself the hero…but as a community manager, I should know that you’ll get a lot farther making someone else feel special!