Sue Halpern writes in the New York Review of Books, giving another explanation for Trump's win, poiting to fake followers of Trump on Twitter:

Andrew McGill pointed to an analysis of five hundred pro-Trump Twitter accounts that had encouraged voters to lodge complaints with the FCC about the Cruz campaign, the majority of which had previously tweeted “17 Marketing Tips for B2B websites.” In other words, they were fake supporters bought and deployed to push a message and look like a small army of concerned citizens while doing so. According to the website Twitter Audit, 4,645,254 of Donald Trump’s 11,972,303 Twitter followers—about 39 percent—were bots, compared to 524,141 of Hillary Clinton’s 10,696,761, or just 5 percent. Here was another way that Trump triumphed.

After studying four million election-related tweets created between September 16 and October 21, the University of Southern California computer science professor, Emilio Ferrara, and his colleagues, determined that one in five were generated by bots. And once they were, they were retweeted again and again by actual humans, who sent them ricocheting around the web, especially those that were antagonistic; in earlier work, Ferrara’s group found that negative tweets traveled 2.5 times faster than positive ones.  “As a result, [the bots] were able to build significant influence, collecting large numbers of followers and having their tweets retweeted by thousands of humans,” and leading to the “spreading of content that is often defamatory or based on unsupported or even false, claims.” Ferrara further noted that, “previous studies showed that this systematic bias alters public perception. Specifically, it creates the false impression that there is grassroots, positive, sustained support for a certain candidate.”