Isn't the problem half about recruitment?

Yes, the numbers do suggest that recruitment is a problem. The number of New Wikipedians joining per month has been falling, which suggests that either fewer people are trying to make their first edit, fewer people are making it to the 10th edit, or both. Fixing the recruitment problem without fixing the retention problem is likely going to result in a lot of wasted effort. We could drive twice as many people into the editing experience, but if they all have the same difficulties (e.g., problems with wikitext, issues with reversions, etc.), then we may just burn through a lot of people who might have become valued editors under different circumstances. I personally agree with the notion that Wikipedia is in competition with other websites for people's time. At a minimum, these companies set user expectations for how a contribution-based website should behave (e.g., rich text editing, user pages, emails when relevant things happen, etc.).

I also just posted the data behind the graphs in case you're interested.

Howief01:06, 18 March 2011

I have to agree with you. Solving the retention problem without improving the new user experience is likely to lead to many more frustrated people, even if it boosts our numbers just by putting more people through the "hazing process". But I don't think this is an either-or. There are probably a few very simple things we can do to boost recruitment, which would multiply the benefits of a better new user experience.

Thanks for posting the data! Hope things are going well.

Randomran02:32, 18 March 2011