In this blog post, Fred Wilson says:

"Let's remember one of the cardinal rules of social media. Out of 100 people, 1% will create the content, 10% will curate the content, and the other 90% will simply consume it. That plays out on this blog, that plays out in Twitter, and that plays out in most of the services we are invested in."

The contrarian in me wants to dismiss this "rule." The right service just hasn't been built yet!

Is that naive? Will there ever be a platform that compels 25% of its users to actively creating content?

I've never bought into this. I do buy that more people will always consume than create—a lot more—*most* of the time. But the ratio is highly dependent on the form of content and the shape of the network.

Some media is just more casual to create than others (even if you limit it to "social media"). The more casual the media type, the higher percentage of people who will create and the less-likely those creations are to be widely consumed.

Example: The percentage of people who create content on Instagram is way more than 1%—I wouldn't be surprised if it's over 25%.

Contrast that to YouTube: I'm guessing way *less* than 1% of users create content (at least on a regular basis).

I'm speculating in both cases. Anyone have numbers?

1st it's important to remember that the 1-9-90 rule is product, not individual, specific. That's to say, you might be a creator on one service, but only a curator or passive consumer on another. It's not that 1% of the population are creators.

Now w that caveat, i still think 1-9-90 is a product failure not an equilibrium, with the biggest opportunity to move folks from the 90% camp into the 9% bucket.

On YouTube that means commenting, making playlists, etc even more than stressing uploading. [cont]

I first started playing with this in 2001 when designing Second Life. My pyramid was creators, editors, personalizers. In that environment you could imagine creators liked to build from scratch, editors like to mix and match (assemble from a model kit so to speak), and personalizers liked to change the color of an object and put there name on it. We found that if we could get people to start connecting with the virtual environment by taking some sort of activity that made it feel like "there's," they were much more likely to come back a second time.

As contributing to the system with simple gestures (repinning for example) becomes easier, we'll see something more like a bell curve perhaps than a pyramid.

Anyway, i think designers should strive for the satisfaction of creating while minimizing the work involved - that's what ignites the 90%

"i think designers should strive for the satisfaction of creating while minimizing the work involved"

Sure...although, what I see this meaning a lot these days is the bar is very low for "creating" something and getting that satisfaction, but the ceiling is very low, as well. Meaning, there are whole services of "easy to create" stuff, which has low community value.

I think designers should strive to let people contribute very easily and then make better and better things if they're willing to put in the effort. Even better: Let people create even better things together.

Wikipedia is a great example of this, but there should be many more.

@ev. great points.

let me suggest that many of these "creation" environments don't get bigger because they don't have economic systems.

I don't believe you can use just economic incentives to create healthy creator-driven products - you need tools and community first. But those which fail to then help creators get paid directly or indirectly will remain the domain of hobbyists. That's the assumption i've bet my career on:

Second Life: here are the tools to build a world, here's the community to do it together, w a real economic system

AdSense: anyone can be a web publisher, build a community and we'll take care of sending you a check

YouTube: anyone can be a broadcaster, views are the currency, exchange your views for ad dollars

As you said, Ev, the more you lower the bar so goes the ceiling.

The products in the top right of this graph (h/t @eladgil) are great, I use them all. But you might as well relabel the bottom of the Y-Axis, "pure keystrokes," and the top, "pure clicks." And a service can only be so additive if the user is barely typing.

Will the bottom left of this graph ever feature a service that boasts something closer to a 5/15/80 ratio? Or 10/30/70?

In order to do that, I think the product would need to make their users feel like better writers, in the same way that Instagram makes me feel like a better photographer. I wonder how you could do that? I know Dustin Curtis is hiring a free copyeditor for Svbtle users.

If Branch had a Quick Edit feature I would amend this statement: "And a service can only be so additive if the user is barely typing."

You barely type in Twitter and it has changed the world many times over. Hopefully, you get the point I was trying to make.

So 5/15/80 in the bottom left of this graph? Is it possible?

Putting aside the quality question, I think mobile ubiquity and other more seamless means of content creation will push the percentage of creators up.

The biggest reason I'm ambitious about Google Glass (and other cyborgian initiatives) is for their potential to make content creation even more frictionless.

@ev is right. some services break this rule. instagram and facebook come to mind. but i like this rule because it reminds me not to expect everyone to create content and you need to invest in the curation and consumption experiences because that's where most of the users will be

The focus on simpler creation is spot on, but "enabling" quality is more difficult because it's in the eye of the beholder. Accordingly, people want to know that their content is well-received, giving them confidence in creating additional content. Lightweight social feedback can really help break the 90/9/1 rule by given content creators positive feedback/encouragement and by turning consumers into creators via simple mechanisms for providing feedback. Path smileys are a fairly good example of this.

Interesting! We purposefully left a Like/Favorite-equivalent out of Branch. Our concern was that it would [negatively] impact the content created because users would optimize for reader approval.

Do you think that's a mistake? Will less content be created on Branch because users aren't getting virtual pats on the back?

In the case of Branch, the creator is getting positive reinforcement prior to the creation of content by virtue of being invited into the conversation. The user is "expected" to add to the conversation, which is a compelling form of what I described. I do think that some lightweight feedback from consumers who are not in the conversation could be interesting to test within Branch. That could start with something as simple as collecting data passively, like views, and sharing it with the content creator.

Late to the party, but let me say that, yes, you can easily break the 1/9/90 rule, but generally only in ways that make the product or service worse.

What we're talking about, when we're talking about the Horowitz Triangle of 1/9/90 (or 80/20 or any of the other classic signature proportions), is a powerlaw distribution (hpl.hp.com). This is a distribution whose metric *is entirely internal to the system in question.* If you are ranking Wikipedia editors, for example, or Linux contributors, then 100% of your sample is contributors, by definition, contributors. Yet even here, the modal frequency of contribution is 1 edit or patch, while the most active contributors are thousands of times more active than the modal user.

Sometimes, when people want to escape the rule, they mean they want high modal participation. This is trivially easy -- require participation. This is like the all-or-nothing logic of newspaper paywalls, with the same effect -- most lurkers will abandon ship. It gives you a 100/0 distribution, but is cataclysmic for headcount, and destroys the pool of likely active users.

A lot of the time, though, what people really want is to restore the happy situation that existed in the 20th century, where almost no one knew _how much_ people cared about a product or service. The most passionate viewer of, say, Star Trek, and the person who could just barely stand it, were, from the point of view of Nielsen, the same *kind* of person.

Now, though, when we see services where the most active user is 1000s of times more active then the majority of lurkers, it hurts our feelings. "Why don't those damn lurkers come out and play?!" Like a woman with a bad boyfriend, the perceived slights of the people who don't like us weigh in so much more heavily than the people committing their free time to participating.

My advice is to assume, then ignore, the 80/20 distribution. (Long-tailed distributions are as close to a natural law of internet use as we have). Concentrate, instead, on making something where participants generate value for themselves and others. The resulting powerlaw will be entirely epiphenomenal, but, when it appears, can be taken as a sign of a healthy service.