I think this guess from /r/typography/ is worthy to consider:
"My educated guess is that it most likely scales in Webkit better. I can only imagine this change was the result of a dozen round table discussions and half a year of a/b testing.
EDIT: Yup! Here's both logos at 50%, scaling done by Chrome. Notice how the letter l changes [1]
It's not a bad theory, but then again the `G` doesn't visually change, the `l` looks better horizontally, but worse vertically. Seems like a tie, at least at this resolution.
Comments
I think this guess from /r/typography/ is worthy to consider:
"My educated guess is that it most likely scales in Webkit better. I can only imagine this change was the result of a dozen round table discussions and half a year of a/b testing. EDIT: Yup! Here's both logos at 50%, scaling done by Chrome. Notice how the letter l changes [1]
[1]: http://i.imgur.com/x6K1yVe.png "
http://www.reddit.com/r/typography/comments/26fgq6/google_up...
It's not a bad theory, but then again the `G` doesn't visually change, the `l` looks better horizontally, but worse vertically. Seems like a tie, at least at this resolution.
Assume google.com serves 300 trillion impressions a year and clicking an ad generates $0.20 in revenue.
The new logo would need to improve CTRs by 0.0000017% to generate a million dollars per year in new revenue.
I really hope not...
A company is making too much money if they can waste half a year in discussing and testing something like this.
More likely they've been working on the scaling algorithm in Chrome, and the logo just happened to look bad (or shall we say less than optimal).
Not really. It takes very few resources to run an A/B test for six months. Set it up, check back for results six months later.
Not that they necessarily did that here, but still.