Filed under: chain-gang,fly-by-wire,Glory, glory, glory,popular demand,wikipedia
Wikimedia had a terribly successful fundraising campaign ths year, with a team of stats-loving traffic and feedback analysts learning a lot about our reading audience and how to connect with them. There was diverse support for the idea of running some banners to promote donating time and expertise and edits as well as money, and some general-purpose “discover Wikimedia” banners were run the first week of January, but this was soon overtaken by preparations for the (wickedly fun) 10th anniversary celebrations.
We should do more of this. The idea of inviting people more explicitly to edit, and running campaigns dedicated to this, is more fundamental to the nature of Wikipedia than fundraising itself. We should be thinking about all year round, spending as much time and effort campaigning for meaningful content contributions as we have for funds.
What would that look like? Here is one idea: WikiProjects could be encouraged to write copy for their own banners, from a hook to a detailed call for what they need. These would be run for a % of new visitors proportional to the project’s capacity to absorb new contributors. A few generic projects would be geared up for a larger influx of editors, and established editors would be asked to help work with those newbies (and to set up comfort zones where they can find and help one another).
The generic projects would ramp up slowly; with one month’s newbies helping welcome those who came the next month. Some new policies regarding working with newbies would need to be proposed on the major wikis, possibly with a group like the original Fire Brigade dedicated to helping the ambassadors and welcomers with the extra load. And the specific WikiProjects could continue to draw in as many new editors as they want, and could try out different messages to attract just the right sort of reader (including efforts at targetting specific kinds of readers).
What do you think? How would you reach out to readers if you could change the way the site looks? (What ever happened to the idea of highlighting the “edit this page” tab?) Over 1% of people who saw the best fundraising messages clicked through them — imagine what we could do if we showed all of those people that they could really edit.
The contribution campaign did run for a little while, with a landing page pointing people toward either a tutorial for getting started, or a list of fairly easy backlog tasks.
But I think the big challenge is identifying really appealing things to do to get people started. Pointing to them to a big backlog of unreferenced or poorly formatted articles isn *not* it.
Comment by Sage 02.06.11 @ 10:06 amSage – What stats do we have on clickthrough rates for those banners, and impact on those landing pages?
I agree that we need to be clever about trying different appealing things for people to get started. If we make it possible for any WikiProject to show a banner to a few % of visitors on a given day, we can see what successful ideas arise. I’m sure many of them will put some energy into the idea.
Some general things we can do are to make the 60-second entry process shorter and simpler (fewer page-refreshes, say — kickstarter has a great pageflow), to make the next 5 minutes of exploration fun, and to add both milestones for account-development and a way to track projects that you’ve dropped into your “interest” list… including the project whose banner got you to sign up in the first place.
Early implementations of these could be static subst:s onto your talk page or userpage, but there are plenty of great examples of sites that make this introduction and progress-tracking fun and natural. OSQA, for instance 🙂
Comment by metasj 02.06.11 @ 1:03 pmWell, for starters, put the “edit” link immediately to the right of the section title, not all the way to the far right. Shouldn’t be hard to do – there are a number of other languages that already do it this way. And it might help to make it more of a button in appearance, to encourage readers to actually click on it.
Comment by John Broughton 02.06.11 @ 9:20 pmSJ,
In regards to clickthrough rates, we have no idea. Currently, without the link going to a landing page on wikimediafoundation.org, there is no way to count clickthrough rates on any banners. I ran into this issue with the 10th anniversary banners, which is why we didn’t do more versions and A/B test them.
Right now the plan is to fix it through getting Open Web Analytics up and running, though according to the latest engineering update we’re still in the planning stages.
Comment by Steven Walling 02.07.11 @ 3:02 amThere are no stats, apparently because the landing pages weren’t on wikimediafoundation.org and so couldn’t do stat tracking.
I like the idea of doing WikiProject-specific banners… especially if they could be tied to which categories the article you are viewing is in, so that you are more likely to see a WikiProject of interest to you based on which page you’re looking at.
Comment by Sage 02.07.11 @ 9:45 pmI don’t think project specific banners will sell, but I do like the idea of periodic ‘discover wikipedia days’. They could be the last friday of each month, for example. I think that is about the right balance of people feeling like the encyclopedia ‘just is’ and recognizing that they can change the state in which they and others find it. To be honest, most people’s hesitancy to edit is partly a phobia of realizing that people as unauthoritative as themselves have been producing a resource upon which they’ve come to rely. Not everyone wants to look behind the curtain. But we should make some effort, even a quarterly, or even yearly! new editor day. Maybe that’s where to start–new editor day–a once a year bonanza of tutorials, collaborations, workshops, etc. Still, much as I want that, until we have WYSIWYG and LiquidThreads, it’s almost unfair to try and drag people into the markup mess.
Comment by Ocaasi 02.08.11 @ 1:10 pmSage – Category-local banners, similar to geolocal banners, is a brilliant idea. Simply making that possible opens up a world of opportunity — and we have basically all of the infrastructure for it already in place.
I would guess that just as having geolocal banners makes distributed meetups vastly more diverse (since you actually get people who are in the area and just saw a banner about the event 5 hours ago), cat-local banners would have a similar effect on keeping WikiProjects fresh.
And, as we’ve seen with the WikiProject assessment system, that’s one of the most scalable ways to cover all articles (or, presumably, reach all of our long-tail readers; the 95% who don’t visit the main page / search page / watchlists).
Comment by metasj 02.17.11 @ 10:53 pm