Illustration of Derek Powazek by Adam Ellis

Did I mention?

I started a farm. Yes, really.

Current Status

I am currently Somewhere In Between.

Talking Plants with Gayla Trail

What'cha Growin?

It was a great pleasure to have this conversation with my longtime garden hero, Gayla Trail. We talk desert plants, my San Francisco cloud forest, and botanical tattoos. If that sound like fun, you’re my kinda plant nerd. Go give it a listen.

What We Can Learn from the Twitter Block Fiasco

An edited version of this article first appeared in Wired.

Here’s what happened in plain English. Twitter has a feature called “Block.” If I block you on Twitter, you can no longer interact with me (follow, retweet, favorite, etc). Last week, Twitter changed block so that previously blocked people could once again interact with the people who had blocked them. And those people were pissed.

After an intense night of thrash, including a lively #RestoreTheBlock hashtag, Twitter reverted the change. Now, then. What have we learned? And what can other companies with large communities take away from this teachable moment?

1. You can’t change the rules in the middle of the game.

People were upset for good reason. They’d spent years adding jerks, haters, and abusers to a carefully crafted block list. The agreement these people had with Twitter was simple: “I don’t want these people to interact with me again.”

Twitter changed that without warning. Suddenly everyone in that blocked group could interact with you again – follow you, retweet your tweets, etc. This broke Twitter’s agreement with its users.

Worse, when those blocked people now interacted with you, Twitter would hide it from you. So if someone was harassing you, all your friends could see it, but you could not. They changed block from something you did to others, to something you did to yourself.

Had they added a new feature called “Mute” that operated the same way the new block did, alongside the current block, no one would have complained. But to change the way an existing community tool worked, especially one so central to the safety and sanity of users, is to court disaster.

2. You can’t make changes in secret.

The change itself was a bad idea. Compounding the badness was how the change was communicated to users: it wasn’t.

Implementing the change with no word or warning created the appearance of Twitter trying to sneak something past its users. I’m not saying that was their intent, but the way it rolled out was guaranteed to make it look sneaky.

Talking to TechCrunch and updating a FAQ is not communicating with your users. Twitter is a communication platform. They have absolutely no excuse for a lack of communication.

When you make a change to a core piece of community functionality, you have to tell your users about it directly, in advance, solicit their feedback, and then listen. You can’t make a change and hope nobody notices. They always notice.

3. Never punish the victims.

The most outrageous part of this whole story is that, despite Twitter’s claims to the contrary, they were fixing the problems of the harassers who got blocked, not the people who were doing the blocking.

I’ve managed many communities with tools like this, and worked with many more, and you’d be amazed how common it is for people who misbehave and get punished to come crying to the admins. But scratch the surface of their complaints and, nine times out of ten, they were the ones in the wrong. “Well, yeah, I did break the rules, but I still shouldn’t be punished!”

Twitter, when faced with years of complaints from the blocked, capitulated to them by restoring access to the users who had blocked them. Then they had the gall to claim that this change would be good for the people who did the blocking.

When making a core policy change like this, always ask yourself, who’s problems are we solving? And what additional problems might we create? Had Twitter spoken to any real users about this change, they would have heard loud and clear what a bad idea it was.

4. Large communities need more than one management tool.

Block is a necessary tool for communities to manage themselves, but it’s not the only tool you need in a community of millions. In the past I’ve likened it to “setting the dinner table with only chainsaws.” We need more humane, nuanced tools for people to manage their networks, and their attention.

What Twitter tried to do is change Block to Mute. A block tool severs the ties between two users. But a mute tool is a simple filter: “Don’t show me stuff from that guy, or stuff that contains these words.” The fact that many third-party Twitter clients have built their own mute tools is proof that it’s needed. Twitter should add it natively without taking away the existing block tool.

There are many more tools that could and should be added. Humans are weird and messy and we require weird and messy tools. Here are a few top-of-my-head examples.

  • People often block because there’s no other way to remove a tweet from view. If Twitter added a “Dismiss” tool for any tweet (just like the one that exists for ads), there’d be less need to block right away.
  • I’ve often wanted a “Timeout” tool that operates like block with a time limit – blocking another user for a day or a two and then unblocking. This would be a great way to short-circuit an argument without implementing a permanent block.
  • I also frequently use a technique I call a “Spike” – blocking someone, which forces them to unfollow me, and then unblocking them again so there’s no trace. This is a workaround – I’d rather just have the ability to remove myself from someone else’s follow list without notification. I see this like leaving a party without saying goodbye.
  • At the same time, Twitter should be monitoring accounts that get blocked or receive other negative feedback for sudden jumps in activity. If a user gets carried away, goes of the deep end, or misses their meds, Twitter could step in. “Hey, we’ve noticed you’re getting a lot of negative feedback today. Are you okay?” Sometimes just getting noticed is enough to calm someone down.
  • Block has always been half-finished on Twitter. Where is the page that shows all the people I’m blocking? There isn’t one. How can a user send a polite appeal to get unblocked? They can’t. Twitter should finish building the feature before changing its core functionality.

I acknowledge that none of this is easy. But neither is creating a realtime multi-platform messaging platform that is used by millions, and Twitter has accomplished that. I know they could do better if they wanted to, and I know that if they don’t, their users will let them know about it. In the meantime, it’s a teachable moment for the rest of us building community online today. One that I hope the rest of the industry is paying attention to.

The Argument Machine

Say you’re a supervillian. Your goal is not to take over the world, but to create more unpleasantness. So you set out to create a device that would ensnare normal, rational people and turn them into ranting lunatics. What would your Argument Machine look like? How would it work?

(Note that I mean “argument” not in terms of academic debate, which is healthy, but in the everyday “unproductive, unpleasant, angry fight” sense of the word.)

First, the machine would have to be online, so it could reach the maximum amount of people with the least amount of construction. People feel protected online, like they do in cars, so they’re more willing to fight. But it shouldn’t be just a website. It would also have to be an app on all the major platforms for maximum reach.

All arguments have a beginning, so we’ll create a tool for people to post their thoughts. This should be as easy as possible, so people do it without thinking too deeply. Forethought is the enemy of a nasty argument.

Arguments tend to fizzle out quickly if the participants are able to make a fully fleshed-out thought, so we’ll limit every thought to a ridiculously small size, say 140 characters. Brevity may be the soul of wit, but it’s also the root of misunderstandings.

We’ll then take those short thoughts and present them immediately to the audience, without any ability to edit them after the fact. We’ll even create a “reply” affordance that seems like it’s private, but we’ll actually show the reply publicly. Nothing creates a fight faster than in-group language overheard by the out-group.

To maximize negative reactions, we’ll design the presentation of these short thoughts to create annoyance. Tests have shown that information coming in too fast can create anger and defensiveness, just like someone interrupting you in real life, so we’ll make everything as fast and interruptive as possible. We’ll also pack the app with lots of notifications, so we can make people’s pockets buzz every time someone mentions them. Just like the electric shocks we used on those lab rats.

If someone isn’t receiving enough of these short thoughts to overwhelm them, then we’ll blanket them with suggestions of other people they should be hearing from, to encourage all new members to quickly reach an overwhelmed state.

With enough people, and enough short thoughts, arguments are sure to occur. When they do, we’ll add heat to them by making sure everyone can see the individual thoughts outside of the argument’s participants. Nothing like a hooting crowd to make a bad situation worse. We’ll even add a tool to allow others to repeat the thoughts to a wider network (out of context, of course). The more people who pile in, the longer the argument will last, the more people will participate.

If everything goes well, the mainstream media will start monitoring these thoughts, so that they can embarrass famous people when they speak out of turn. We’ll even create an way for people to embed these thoughts on other sites so that they remain even after the original poster deletes them. Gotcha!

Nothing is better at creating new arguments than other arguments, so once we reach a critical mass, avoiding arguments will be impossible. Now all we have to do is sit back and practice our supervillian laugh. And maybe go public.


I think you get my point by now. Yes, I’m talking about Twitter.

I’m not saying that Twitter was designed to create arguments. I’m just saying that, if you set out to create an Argument Machine, it’d come out looking a lot like Twitter.

I think Twitter is awesome. As of today, I’ve tweeted more than 16,000 times. It’s the community tool I use more than any other. But with Twitter preparing to IPO and monetize all of us, I worry about the health of the community the tool produces.

If Twitter cared about avoiding arguments, there are so many things they could do: remove the outdated character limit, let us edit tweets, create progressive circles of privacy, don’t let retweets out of our networks, slow the whole thing down, and encourage smaller communities. But they’re not going to do any of those things because the financial pressure points in exactly the opposite direction: more users, more tweets, faster and faster. Who cares if they’re fighting?

Well, I do. I want Twitter to succeed. But every time I see some innocuous tweet spawn another long, bile-filled thread of awfulness, I become less interested in tweeting. I know, arguments will always happen, online and off. But the design of the tools we use can encourage or inhibit this human tendency, and I’m afraid Twitter is falling very far into the “encourage argument” side of things. And I’m not alone.

Where will we go when we all tire of getting yelled at on Twitter? What would a tool designed to create a harmonious exchange of interests and ideas even look like? I’m confident we’ll find out eventually, but I seriously doubt its logo will be a little blue bird.


This post also appears on Medium.

You made it to the bottom! You must be very good looking. Wanna go back to the top now?

Hi, I’m Derek. I make awesome community-centric web stuff. I sometimes post things to Flickr and Twitter. I’m mostly harmless. More.

This site is powered by WordPress and expertly hosted by Media Temple.