I am frequently very annoyed by Google:

I google something, and then I want to view Images of that thing, so I click the second button under the search bar out of muscle memory... except often it can be Shopping, or Videos, or something other than "Images".

Why does Google think it's a good idea to "randomly" rearrange their interface like this?

I just wonder if anyone on the planet appreciates this UX choice, or if there are business incentives to doing it and trapping users on pages they didn't expect to land on.

  • I don't think Google does anything in a random way, the "mistake" could cost millions of dollars
    – Devin
    Commented Oct 4, 2019 at 18:50
  • @Devin the order of the buttons is different for different google searches
    – minseong
    Commented Oct 4, 2019 at 20:35
  • so you hit an A/B test. Again, nothing is random at Google or any other giant company
    – Devin
    Commented Oct 4, 2019 at 21:20
  • @Devin the way I used "random" I meant it as "inconsistent", not randomly-generated. Of course Google has a reason for choosing their button order. I am just wondering what the reasons are for making that order different every time.
    – minseong
    Commented Oct 4, 2019 at 22:20
  • 1
    We don't know because we don't work for Google. Commented Mar 3, 2020 at 8:29

1 Answer 1


In short, no.

I'm answering this in general and not to your specific example.

Consistency is a key aspect in making an interface predictable and easy to use. Moving elements around on the page can result in confused users, therefore typically you should aim to keep key functions in the same places. Your example clearly illustrates this.

However in some cases it can be that the context of the page suggests certain features should have a higher priority and therefore would need a different, more prominent position in the interface. In such cases the designer has to choose between prominently placement for priority features or consistency across the entire application.

Carefully considering the success criteria, a/b testing can be used to come to a conclusion on which decision to make.

Not the answer you're looking for? Browse other questions tagged or ask your own question.