On Wed, Jan 22, 2014 at 12:37 PM, Brandon Harris <bharris@wikimedia.org> wrote:
> I think the main edit button at the top should not be the quiet state from mw-ui. There is very little use of color on the skin now, so it's not like we're overwhelming things. We need to stop being afraid to make a few things LOUDER, particularly since in this prototype we're removing some color (like the Vector borders).So, I tried it with the main edit being blue and it was *crazy* loud. The most-important thing in the page loud. So I made it quiet and it looked better.
It is not that loud, as you can see at http://i.imgur.com/X4uhU2T.png and it _is_ the most important call to action on the page.Now, this speaks to the larger problem we’re trying to solve: user awareness of the edit button.
We know that Vector’s tab layout has trained people to pretty much ignore the tabs, so they don’t see the edit tab (tab blindness, call it). By pulling the tabs into the main content area, we are working to solve that - to bind the actions directly to the content they affect.
Currently the tab blindness causes people to not know they can edit, so we have fewer editors.
I think think that making the edit link loud will definitely *increase* overall edits (in fact, I’ll stake my job on it), but I think it will result in a radical increase of *bad* edits as well. Too many “kick the tires” vandalisms, maybe.
It’s definitely something we can test, though.In the last GettingStarted A/B test, we delivered an "Edit this page" CTA that was twice as large and delivered in a modal,[1] and it did not result in a radical increase in bad edits. There's no need to A/B test this or be scared about bad edits.