When I onboard a new designer, I usually talk with them about Craig's List, Amazon versus "beautiful" UI. The point is exactly the last line of this article:
"Focus on content and functionality when you’re designing new products; that’s the validation that will build a business."
Now I'm going to add this article to required reading for new engineering team members.
Amazon is bad UI though. The information they present is inconsistent, the descriptions barely make sense, it's hard to compare products. It's byzantine.
I Amazon really bad? Their transaction volume would seem to indicate that users are succeeding in placing orders at a very high frequency. This is exactly why Amazon is a great study for new designers and engineers.
Yes, you are right, and I should have qualified that because I'm aware it's optimized, but it's effectively a grey pattern at best: a UI that is sub-optimal in terms of presentation of information, but optimized for click-through.
I'm sure one could make a case for 'if they buy it it was good' but I suggest people are easily manipulated, especially when confronted with a massive machine with $Billions to figure how to.
But even with both of those things said - I think it is still actually crappy because 1) the click through may be optimized and yet some information could be available to make it better that may not affect click through and 2) the 'long term funnel' may be affected by maligned customers; I avoid Amazon, I suggest others might as well though it may not be that many.
Both are important but look is always second to reliability and functionality. You can have the prettiest book cover in the world but if it's a sandwiching a pile of crap it won't sell.
"Focus on content and functionality when you’re designing new products; that’s the validation that will build a business."
Now I'm going to add this article to required reading for new engineering team members.