a feminist technology blog

What percent of users contribute content?

Someone asked me to look this up, so I did a mini-research project. For a while I’ve assumed that online content contribution followed the 80/20 rule (the Pareto Principle) which maintains that for many things, 80% of the consequences stem from 20% of the causes– e.g. 20% of volunteers do 80% of the work.

However, for a while I’ve been suspicious of the Pew Internet & American Life project that 44% of internet users, and 57% of teen internet users, have contributed content. This is a totally unuseful number, since it makes no distinction between Joe Doe who writes a single IMDB review, and Jane Doe who uploads 10 YouTube vlogs a week. You can’t make product decisions on such a vague number.

I started with Jakob Nielsen, who claims that:

In most online communities, 90% of users are lurkers who never contribute, 9% of users contribute a little, and 1% of users account for almost all the action.

Does this hold up after running some numbers? Well, first of all, in order to determine what percentage of site visitors contribute content, we have to know the number of site visitors. I am far from convinced that most sites’ proprietary user stats are correct, as it’s in every site’s best interest to overstate their numbers, and MySpace, Yahoo, Wikipedia etc. don’t tend to be extremely forthcoming about whether they are talking about unique visitors, repeat visitors, user accounts, etc.

If we fiat that all site visitor statistics are correct, then Nielsen’s law appears to be accurate.

Wikipedia: Only about 1-2% of total site visitors contribute content, according to Yochai Benkler’s The Wealth of Networks. (As of 2005. See graph).

I think this number is underclaimed, but it’s true that there are a core number of Wiki contributors who do the bulk of the work of the site. Jimmy Wales has claimed that 10% of users make 80% of all edits, 5% of users make 66% of edits and 2.5% of users make half of all edits. (Source: [ppt, see Slide 25]).

– Youtube: YouTube is currently serving 100 million videos per day, with more than 65,000 videos being uploaded daily. (Numbers from YouTube). 65K is about 0.1% of 100 million. Of course, the number of videos being watched is not the same thing as the number of users. Even so, YouTube has approx 20M unique users per month. 65K * 30 = is 1,950,000 which is 9.75% of 20 million. But, again, we have no idea how many of these videos are uploaded by the same people.

Bradley Horowitz at Yahoo wrote a blog post called “Creators, Synthesizers and Consumers” in which he claims that about 1% of users create content, 10% “synthesize” content (add comments, etc.) and 100% of people benefit.

So let’s assume that he and Jakob are right: 1% hardcore contributors, 10% overall contributors and 90% total lurkers.

What impact does this have for feature design?

1. You don’t want all your users to contribute. The undereditorialization of YouTube already makes it difficult to find stuff you think is really good. If every user was spamming every social software site with mediocre content, the sites would lose value.

2. Not every user is created equal. Keep your contributors happy. Even if they are demanding, picky nerds.

3. On the other hand, contributors are only 10% of your overall users. Make your uploading and editing tools good, but make the browsing/searching stuff really really good.

I’d appreciate if anyone has any stats they could share to prove/disprove these numbers.

Technorati Tags: , , ,


Link Roundup for 11/28/06


How to Embed YouTube Videos in PowerPoint


  1. Interesting numbers and I certainly don’t have any better numbers to prove/disprove those numbers. Just one thought though – the numbers seem to stem from a community, forum, service or whatever – I am wondering to which degree people produce materials for other contexts e.g. how students may rework some Wikipedia citations into a report, use a number of pictures from different sites to produce a slide-show for some friends etc. etc.

    So I am wondering if we can use this number as a general rule across domains (they certainly seem to hold through for individual domains) – so though people are not ‘producers’ on Youtube they might use Youtube as a source for being productive in other contexts…but probably this is somewhat difficult to account for…

  2. I’ve looked fairly hard at how people use Ryze. About 5% of the people are very active and attract about 60% of all the homepage hits. Another 10% are occasionally active and attract about 30% of all homepage hits. That leaves 85% of all ryze members with about 10% of the attention. However most of that 85% of Ryze members never or seldom come to the site.

    I’ve noticed significant development in the knowledge, skills and confidence of people who use Ryze regularly. I’ve seen people develop from being unsure of themselves and unwilling to venture an opinion about anything, to becoming more vocal, and finally to taking up significant leadership roles.

    This is entirely because they read other people’s comments and they eventually engage in the conversation. They join groups as uncertain spectators, they become full members and in the process they take on the knowledge of the group.

    That facility is available for all Ryze members. Bur only about 5% of members have that experience. Only 5% of members put in the work. You get what you give, as they say.


  3. Thomas, interesting point. I think it’s almost impossible to say “how many people produce things that in some way use the internet”, since that’s a really difficult metric to measure. I’m not sure also how useful that information would be. I can say, for example, that the vast majority of *my* students use the internet for research. But what’s the difference between, say, reading the weather report, and reading a Wikipedia page?

    I’m really interested in user-contributed content sites and communities, but you are quite right that user practice doesn’t exist within specific sites – it moves across them. When i put a powerpoint together, I use pictures from Google Images, videos I rip from YouTube, free templates I get from PPT sites, etc. etc. I don’t share the PPTs after I finish them, but I am still *creating* something.

    I guess I’m interested in interrogating all these claims of democracy, etc. that presume enormous amounts of contributors.

  4. John, this is really interesting data. The process you describe – people gradually being socialized into a community of practice – holds true for participating in offline communities, or indeed any sort of learning situation, really.

    So the entire Ryze community benefits from the 5% of users who do most of the work.. but that community consists of only about 15% of the overall users. I wonder how that compares to the numbers at other sites.

  5. interesting, I made similar observation! The ratio is very low and I am concerned that the norm for non-contribution might increase. I do think our web 2.0 can be better and smarter if more will think they are critical. I try to dedicate my own contribution and efforts to this issue. Your insights are correct and match other “public good” research data (outside the net). I find it important to do something about it.

  6. I want to get the percentage, but I am not sure which is really accurate!

  7. Hi! I was surfing and found your blog post… nice! I love your blog. :) Cheers! Sandra. R.

  8. I love your site. :) Love design!!! I just came across your blog and wanted to say that Iíve really enjoyed browsing your blog posts. Sign: ndsam

  9. Sign: umsun Hello!!! rcuwwymhyw and 7145ssgfhphzye and 5610Thanks. We look forward to hearing from you again and for your opinions on the world of work.

  10. Very great post. I simply stumbled upon your weblog and wanted to say that I have really loved surfing around your weblog posts. In any case I will be subscribing to your rss feed and I’m hoping you write once more very soon!

Leave a Reply

Powered by WordPress & Theme by Anders Norén