My guess is that voters don't consciously think this through, but rather, they intuitively understand the significance of their decision. This NY Times article (about why Iowa didn't give Obama more momentum) seems to corroborate my suspicion that it takes two wins to create the bandwagon effect:
Contrary to some pre-election analyses, there was no reason to expect that Obama’s Iowa victory would have much effect on New Hampshire voters, according to scientists who study so-called preference cascades. In studies of cascades, people do tend to follow others’ lead against their own preferences — but not until at least two others have made a choice, said Duncan J. Watts, a sociologist at Columbia and senior scientist a Yahoo Research. Mr. Obama needed to take New Hampshire to get the snowball rolling.