4 ms·

> "People you may also know"? Really? One of their objectives is to keep expanding their user base. This feature is crucial to be in a visible place on the scr

by nnq 13y ago

> "People you may also know"? Really? One of their objectives is to keep expanding their user base. This feature is crucial to be in a visible place on the screen. Also quite useful for users that have just joined Facebook. And about: > random stories "Top stories" rather than sorted by date It's a known fact that when you add randomness to the action -> reward relationship, addictiveness significantly increases. (it's not intuitive, but has been proven on everything from rats to humans!) I stopped using Facebook "personally" some time ago, but I admit that they know their shit and do it well. Whether they do it by brute-force metrics or they have some really smart UI/X evil geniuses, I don't know, but it's well done!

voicereasonish 13y ago
That's the thing about tuning based on metrics though. You end up with a product that the majority tolerate, but no one really loves.
- nnq 13y ago
  Yeah, this is what you get if you use metrics to find a "common ground of preference". But this is the primitive way of doing it. We already know the better way to do this, thanks to things like Howard Moskowitz's research and all the work done afterwards (http://www.ted.com/talks/malcolm_gladwell_on_spaghetti_sauce http://www.ted.com/talks/malcolm_gladwell_on_spaghetti_sauce). But nobody has successfully applied this to commercial UI/X design yet. Maybe because people are not sophisticated enough to appreciate choice when it comes to apps/websites. Or maybe because they are so overloaded that making them make any choice actually scares them away and drops the retention/click-through/interaction/etc. numbers. Then again, if would work, the company successfully using this would get a monopoly in their field instead of naturally sharing the user base with other competitors simply because the users don't all like the same things. I kind of like the idea of having 4 different tomato sauces made by 4 different producers, the "letting the market solve the problem" way, so I'm secretly glad the "better way of doing UX/I by numbers" hasn't been well applied in software :)

by nnq 13y ago

voicereasonish 13y ago

That's the thing about tuning based on metrics though. You end up with a product that the majority tolerate, but no one really loves.

nnq 13y ago
Yeah, this is what you get if you use metrics to find a "common ground of preference". But this is the primitive way of doing it. We already know the better way to do this, thanks to things like Howard Moskowitz's research and all the work done afterwards (http://www.ted.com/talks/malcolm_gladwell_on_spaghetti_sauce http://www.ted.com/talks/malcolm_gladwell_on_spaghetti_sauce). But nobody has successfully applied this to commercial UI/X design yet. Maybe because people are not sophisticated enough to appreciate choice when it comes to apps/websites. Or maybe because they are so overloaded that making them make any choice actually scares them away and drops the retention/click-through/interaction/etc. numbers. Then again, if would work, the company successfully using this would get a monopoly in their field instead of naturally sharing the user base with other competitors simply because the users don't all like the same things. I kind of like the idea of having 4 different tomato sauces made by 4 different producers, the "letting the market solve the problem" way, so I'm secretly glad the "better way of doing UX/I by numbers" hasn't been well applied in software :)