Not sure in this situation, but one technique that also gets done is A-B testing on a live site. You subject some subset of the population (A) to your new change and then compare the response to the other population without the change (B). So perhaps it was there by mistake but there are also scenarios where feature trials are intentional.
Well, if it was intentional and I was in the subset - -
they sure got an earful from me.
One time I had a Disney ad take over my phone in the middle of a conversation.
Do they really think you are going to buy when they irritate you like that?
Anyhoo, thanks much. - - J