Interview with: Danny Horn, Wikia

Interview participants: John Fowler and Tyler Thornton

Interview date: 1/15/10

Quick background:

  • Muppet fan before there was an internet
  • Started Muppet Wiki in 2005
  • Fell in love with the platform completely because friends were there and we were doing stuff together; “People are doing stuff and I can see what they’re doing”
  • Now a product manager and Wikia
    • Designs tools like myHome, wikistickies
    • “This needs to be fixed”


Approach to user interface and user experience

  • To get a different kind of user, we knew we needed the user interface to be more user-friendly and less “nerdy”
  • Wikia has a commercial motivation - in order to live we need to sell more ads, and in order to sell more ads the site needs to be more accessible.
  • We want new founders to be more accessible, we want new editors to be more successful – what are the things that are getting in their way?
  • Every wiki is like a little laboratory – so if we want to experiment with stuff, we make a change that only applies to new wikis. Different wikis and have stuff turned on and off so we can see what works.


Tools/Approaches that Wikimedia can learn from:

  • Made type bigger for important buttons
  • New save bar
    • Moves up and down so at bottom of browser and visible no matter what
    • Led to a big increase in save rates
  • WYSIWIG editor
    • Knew people were having a hard time clicking on the edit button and seeing all the code
    • Dramatic increase in save rates – 93.5% for logged in, 79.2% for total, 41.9% for anon
    • Launched WYSIWIG as a switch you can turn on and off. Every existing editor was set to off but every new or anon person got the rich text editor
  • myHome
    • Only for logged in editors
    • Shows activity feed --> name of the page, name of the person, change that was made --> could potentially filter/highlight areas of interest (pages you’ve edited, related categories, stuff on watch list)
  • Hot spots --> pages with most edits (in the last 3 days)
  • Showing edit counts in big numbers on user pages
    • Very transparent how much editing experience people have
    • Adds incentive for users to be active and add to edit count?
  • WikiStickies – “show what we want people to do”
    • Relatively new tool
    • Automated feed of pages with out images, brand new pages for spell check, etc. and wanted pages (with red links going to them). Pointing people to wanted pages has not been successful - better if not a heavy lift?
    • Community members can also now create stickies – admins can say “there are things I really want”
  • Is doing a new little study looking at anons and new logged in people
    • What’s their first edit? If we want more women to be editing, what do we have to do differently? Twilight vs. Star Trek --> are people editing differently?
    • Obvious thing that popped out --> most common kind of first edit is to a long pages and common pages and most common type of edit is half a sentence or adding to a list --> “I see exactly where I can put a little bit more information”
    • Low risk? I know I am doing the right thing


Other suggestions for Wikipedia?

  • Logged in only editing
    • 20 wikis on Wikia do this
    • Found evidence that says that it leads to better content and better community
    • Give the new people names to reduce frustration and backlash against new people --> Think about an in-person volunteer project and 50% of people have masks on
    • No evidence that having people log in actually makes you lose editors. You end up with more active editors when you turn off anon editing than when you don’t

From a follow-up email about the development team:

“We've got 15 full-time engineers working on product stuff and bug fixes, plus a few more people doing Operations. For any given project, there's usually one to three engineers staffed on it. A small project like the save bar was one person for maybe a week... a big one like the Wysiwyg editor was a team of about five people working for a few months”