This is so non-obvious and conditional probability is so hard for people to reason about that I think this needs an example: I want you to think about how many transactions you've had involving a payment at a bank. Thousands, right? Now think about all your calls with banks.
-
-
Show this thread
-
Your call never starts with "Hello, calling to inquire about that transaction I had using my credit card with a diner recently. Everything went perfectly and the sandwich was good. Have a nice day, toodles." And you're rarely thrilled at the end of the call, either.
Show this thread -
This phenomenon has two concrete consequences for business: 1) You're going to be encouraged by substantially everyone to let customers heavily influence your product development but *the customers who talk to you are not necessarily representative.* 2) Hug your CS team.
Show this thread
End of conversation
New conversation -
-
-
And part of why standard sentiment analysis isn't particularly useful for call center use cases.
Thanks. Twitter will use this to make your timeline better. UndoUndo
-
-
-
That's why the #1 quality metric for most business processes to maximize is "what percentage were correctly executed the first time". Focusing only on visible problems can result in whack-a-mole & firefighting.
Thanks. Twitter will use this to make your timeline better. UndoUndo
-
-
-
This is one of the challenges for utilities like telecoms. If you do your job perfectly, they don’t think about you at all. If they have to call you, it’s almost always because something bad happened.
-
Yet they value a quick fix and friendly voice over never breaking at all!
- 4 more replies
New conversation -
Loading seems to be taking a while.
Twitter may be over capacity or experiencing a momentary hiccup. Try again or visit Twitter Status for more information.