Answering your question head-on, you would consider a pivot if the product you have is of no value or does not solve your target market problem. With that point, I agree with other contributors on this forum. If you have validated the problem you are solving prior to the launch, chances are that you wont be "highly" surprised. But realistically, there are always surprises and aha-moments when the product hits the floor and you have live customers with live data using your product. Problems should be anticipated and iterations on your product should be welcomed. Despite it all, if you're product activation and engagement metrics are not where it needs to be, then you should consider a pivot.
But you are entering a learning phase here, prepare your team to be vigilant about product metrics and embrace the insight you will get from your customers. Who knows the next pivot you do (if any) might be highly successful. Check out stories of famous pivots here: http://www.businessinsider.com/most-successful-pivots-2011-4