I am currently working on a redesign and just saw the agency’s suggestions for animation. Seem to be highly influenced by iOS apps’ transitions and motion. But this is a website, for browsers, to be displayed on a variety of screen sizes, on a variety of browser versions.
And the users of my website use mostly other websites so if I create something that behaves very differently, the training they have will be wasted and the experience — unintuitive.
So the agency will present this to high management, in an ideal layout, designed in animation software, moving perfectly smoothly and solving for only one case without a clue on how to approach corner cases. If management finds it sexy, we’ll be stuck with development and maintenance of ultra complicated code. And the pleasure of discovering and resolving all corner cases ourselves, on a deadline, with limited budget.
I don't think even that is a good enough reason. They simply don't approach the problem the way someone using their app would. It's the developers' problem to turn a difficult design into code, but the real problem is that even when that is done perfectly, the UX still sucks.
Every time we redesigned a major UX component of the dashboard of our (desktop) app, we would grab a random secretary, HR person, accountant, or anyone else we could find and borrow an hour of their time and just watch them try to use it without any prior preconceptions that come as a result of designing/coding the interface yourself.
It's been 10 years since we used that trick and I've moved on from that company, but I still think it's the best approach I've ever come across short of outsourcing some sort of panel testing.
Oh lord, I don't really have to deal with complex animations on websites. Even though I wouldn't be the one physically building it, I don't look forward to the day. That sounds like a 2d problem becoming 3d.
I work with web designers and can corroborate that.
Most of the time, their product is sold, negotiated and approved before anyone has actually used it. An unusable mess can be a "successful" project, while a usable product can be completely kiboshed and not even paid for or used. This pushes the emphasis towards beauty instead of functionality, often to the latter's extreme detriment.
This gets worse when the stakeholder has some pet attribute they believe is important, especially "cleanliness." These "clean" designs mindlessly strip away anything that can physically be stripped away, without regard to its importance. As long as the design still kind of reminds the client of their company, they don't see it as a problem.