This is actually such a frequently discussed topic but also a very difficult one. There’s a million experts sand a million answers, and we have no real idea if any of them are right. Neither do the experts themselves. The data is noisy, understanding motivations is hard. I’ve prognosticated on a lot of this stuff myself, and I think the real answer is don’t think about things like this, instead think about process, and build tactics from good process. To me such a strategy would have several guideposts/rules to follow:
1. Campaign on how you can make people’s lives better. I can’t name a single, non-incumbent President who has ever won that didn’t do this effectively. At least not in the modern era in which Presidents personally campaigned. Keep in mind this rule is open to wide gulfs in implementation. Compare the FDR campaign of 1932 to the Trump campaign of 2016. Both actually were campaigns where the candidate vigorously focused on how he would improve people’s lives if they voted for him. Trump’s was a far more negative campaign and focused in part on how he would improve people’s lives by going after the “enemies” of his voters.
2. Do not campaign on issues that don’t directly relate to 1, and especially stay away from issues that most Americans don’t care about. This is probably the biggest actual, strategic level mistake Hillary made. She spent way too much time calling Trump a stupid racist and not enough time on building her own message as to why a vote for Hillary is a vote for better times ahead. Hillary had the actual policy proposals and platform, but politics is more than just “having those”, it’s what you choose to spend ad buys and public appearances on.
3. We live in an era of big data, use it and adapt as it changes. We now know that Hillary’s campaign started to see internally that the Midwest was in trouble midway through the general. There was some classic groupthink going on where so many around her inner circle basically didn’t want to admit they were in trouble. So instead of addressing it at all, they ignored it. How exactly she could’ve responded is up for debate, what we know is about the only response we saw was a very, very late shift to make a few previously unplanned campaign stops.
4. Controversy creates cash. More than ever before we live in a 24 hour news, social media world. Getting attention gives you a stage and gives you a platform. Don’t ignore that once you get a platform you can say whatever you want, but getting a platform requires doing things that are exciting and generate interest.
5. There are 51 elections for President, not one. There is correlations between those elections and national sentiment matters, but you’re fighting for electoral college votes. At all stages of your planning you need to run things through this filter, if your data shows you’re struggling in states you need to win, you need to address that specifically. Don’t get stuck on “well if I do something that helps in Michigan it might hurt Hispanic turnout in Texas”, how are you doing in Texas? Does it matter? Focus on what matters.
6. Turnout and persuasion both probably matter, so you should probably do both. There’s good evidence that Trump was able to persuade some people to his side and that helped him. There’s also evidence that Hillary’s get out the vote process in places like Philadelphia, Detroit, Milwaukee and other important locations was lack luster. This is really campaigning 101, but “turnout strategies” and “persuasion strategies” aren’t either or, most successful campaigns try to do both, and they try to focus them in the states that matter (see point 5.)