I personally prefer the "slow walk" of UI improvements instead of a full-blown UI redesign. Examples are eBay, Paypal, and (in a way) Amazon. eBay and Paypal are both stuck in a half-design stage, where 50% of the pages are designed and 50% of the very functional pages have been redesigned but not rolled out... in perpetual beta, because the redesign cannot accompany all the features of the old. Amazon very slowly rolls out new button designs and side-menus, probably whenever Bezos is sleeping. But some of the buttons in the rarely-used sections of their site are still using the old familiar yellow JPGs.

But what if you have a total of 3000 engineers. What will all of them do? Perhaps a handful of them can investigate and try new things?

The article is based on the assumption that upper management is just trying to throw out new UI for no reason other than to say they did something new.

The reality is these companies change the UI because some metric told them that the new design will bring in the most ad revenue with the least risk. They know most people don't care, and those that do will become apathetic. They don't care about your taste or feelings, just about the percent likelyhood you will click on an ad.

I'm not saying this to be overtly adversarial to these companies, but it is a little naive to think they 1) are just arbitrarily changing things or 2) don't know that people will hate it, but it's a calculated risk.

Thanks, that was a rereshing read. So many companies do these, sometimes I feel like I'm the only person in tech who hates UI redesigns that don't add or fix any functionality. Moving around and changing the appearance of UI elements is worse than useless, it takes users longer to perform the actions that are the actual reason they use the site.