Most of the viewers who tuned in to the program the second night had also watched
Why this is right
As written, we could say that this answer is necessary in order for the author's causal explanation to be plausible. The reason the 2nd night's audience was almost as big as the 1st could only be due to "most of the 1st night's audience liking what they saw" if most of them continued to watch during the 2nd night". Alternatively, we could see if the negation weakens. If we negate this and say, "Most of the people who watched it on the second night did not watch it the first night", then it sounds like Friday was mostly a new set of viewers. This suggests that there must be an Alternate Explanation for the Thu / Fri audience numbers, since the two audiences are mainly different people. (note: it is mathematically relevant that we know the 2nd night had a slightly smaller audience than the 1st night had. If more than 50% of Thursday night's audience had carried over to Friday, then those Thursday-watchers would be more than 50% of Friday's audience too, since Friday's audience was a slightly smaller total.) (other note: the word "most" is wrong on Necessary Assumption, 99% of the time we see it, but it's a totally acceptable word if the conclusion is talking about Most A's).
Skill tested: Necessary Assumption · how this choice captures the argument's function is the move to repeat next time.