The market is literally flooded with Pascal cards, despite the large demand (due to Turing being such a horrible value, and no Vega VII/Navi yet). Maybe they cleared their inventory of chips, but OEMs still surely have cards on their hands.
There is a rather large market disruption happening next year with new AMD arch and Intel entering the game, which is a huge thing and a threat to Nvidia's dominance by the looks of it. I think Nvidia will have to come up with 7nm cards fast as a very quick follow-up to Turing. They need smaller dies with higher performance for cheaper. Otherwise, people will be left with sour tastes in their mouths and go to Intel/AMD in 2020. Surely they will want to enter the market with a bang - that's the only way to introduce new series and recoup costs of introducing it and supporting it in the future. They will want to sell a lot of units.
Nvidia will have to launch a great series right before, so people massively jump onto. When competitors launch their products, most people won't have the need to upgrade then. That's the only way out of this predicament that I can see for Nvidia. From the consumer's perspective, I don't see any reason to get Turing. It's a bad value card that's outright weak - they launched with similar ballpark of performance as 3-year old Pascal cards (which were great then), on an outgoing manufacturing node, and right before a large change to the GPU space that surely will be a great time for the consumers.