A few months ago, I wrote that Donald Trump would win the GOP presidential nomination – but that would be the end of the line for him. I was sure that Trump just couldn't shore up enough of the already too small Republican base to win in November, thus nearly guaranteeing not only a loss but a big loss to Hillary Clinton in the general election.
Well, I was wrong.
In the 80-odd days since I wrote that piece, I've been seeing more and more evidence of why my predictions for Trump's demise were wrong — and that his chances of winning in the general election look pretty decent.
The biggest reason is something very familiar to CNBC's audience: management. In this case, it's the kind of disruptive management that refuses to accept all the conventional wisdom and truly disrupts the status quo.
Trump, a supposed political neophyte, seems to understand the Republican base better than party leaders. Trump may have failed to win over all the conservative elites represented by people like Bill Kristol and the National Review editorial board. But what I and others forgot was that on Election Day, there's not enough of that conservative elite base to fill a phone booth. Securing their support is no way to win a general election.
And, as Mitt Romney found out the hard way in 2012, even being a more moderate mainstream conservative with experience winning and running a liberal state like Massachusetts isn't good enough to win the White House anymore. And it's also impossible to hold down the conservative support and expand the potential Republican voting base at the same time.
What Trump and his advisers clearly realized a long time ago was that it would have to really disrupt the hardened "red/blue" divide to win. America's demographics, news media, and educational establishments have all successfully destroyed the traditional Republican message for at least a generation.
As someone who had never run for office before, Trump already had the best chance to run away from that party title and shrinking base to create a new coalition of voters. But he had another advantage I missed in this area: his bluntness.
To use the kind of blunt language Trump is so fond of, the current Republican Party is a "loser." So, Trump has to regularly prove he's not a part of that losing team while still getting the Republican National Committee's money and ground-game support on Election Day.
So far, that disruptive strategy has worked. He started by successfully capturing the attention of blue collar and union workers with his attacks on open borders and U.S. manufacturers outsourcing to Mexico and China. The coarseness of that message successfully separated him from more cautious Republicans. And by grabbing hold of a populist pro-blue collar message, he defused a traditional Democratic Party weapon. He continued by raising concerns about Muslim immigrants and even Muslim tourists in light of the San Bernardino Jihadist shootings. That blunt talk was considered foolish and even politically suicidal at the time, but it undoubtedly helped spur not only Trump's primary victories but also record participation in the Republican primary process. No Republican has done anything like this since Richard Nixon stole the Democrats' thunder in 1972 by not only withdrawing quickly from Vietnam but also making his historic trip to China.
Base? Trump doesn't need no stinking base.
And it hasn't ended there. Trump is still breaking conventional rules by recently insulting New Mexico Governor Susan Martinez, a GOP "golden child," because all the conventional wisdom says Republicans need more women and Latino voters to have a future. But remember, Trump is trying to make sure you don't primarily identify him as a "team player" Republican anyway. That team is a losing team and Trump wants little part of it. And he's probably also aware that it's a waste of time for any non-Democrat to run after elusive female and Latino voters anyway. It sounds crazy to slam Martinez, but as Trump is proving over and over again, Trump's campaign is crazy like a fox.
It also sounds crazy to a lot of people that Trump has been actively going after the white vote. Why does a non-Democrat ever have to do that? Because white voter turnout has been down in recent elections. Trump knows he needs to energize lots of white voters who have recently stopped voting. He did that in the primaries and it's all still working now.
GOP Chairman Reince Priebus is still coming along nicely and showing more support for Trump day by day. According to the latest NBC/WSJ poll, 86 percent of registered Republicans now support Trump over Clinton, up from 72 percent a month ago. By the end of the GOP convention in July, that number should be close to 95 percent. The remaining 5 percent of Republicans who will never support Trump won't matter. The conventional wisdom about shoring up your base was all wrong.
The second biggest mistake I made about Trump is something else the CNBC audience should appreciate: I didn't think his incredible abilities and experience at self-promotion would translate very well from the business and entertainment media world to the political arena. But I forgot that Trump has been a master business marketer for decades and has also been working closely with some of the best writers in reality TV for more than 15 years. And probably the best talent those writers have is making events and comments sound truly off the cuff and natural even when they are really completely planned and strategically weighed.
I don't think Trump has said one thing or sent out even one tweet during this campaign that didn't sound like something he truly believed and would naturally say or write. Even if you've hated 100 percent of the things Trump has said and written, it's important to understand that Trump has won a crucial marketing and persuasive victory simply by convincing you that what he's saying and writing is his genuine voice and authentic personality. It's called building a clear and identifiable brand. Winning an election is still very much about connecting personally with key voters and you can only do that if you present a clear personality or brand to the voters in the first place. If you're the person who sees Trump's personality/brand and have decided you hate everything about it, I have news for you: You're not the target audience. But you're still proof that Trump's messaging is at least very clear and that's often more than half the battle in business and politics.
And that brings me to my last mistake about Trump's chances: I underestimated how bad Hillary Clinton's campaign would be. To be fair, I never thought Clinton was a particularly strong candidate. But at every essential task of marketing and messaging, the Clinton campaign has been surprisingly bad.
We all know Trump's key slogan/promise is "Make America Great Again." I'm still not sure what Hillary Clinton's key slogan/promise is and I follow her campaign very closely. Is it "I'm with Her?" If so, it's not very good in that it doesn't seem to have anything in it for the person who isn't "her."
Clinton's Twitter feed and website are helping Trump immensely as they seem to pump out phrases like "a Trump presidency," and "President Trump" more often than Trump does himself. These kinds of messages present and reinforce the idea of an actual President Trump in our subconscious brains. This is why the old TV commercials for consumer products when we were growing up used to avoid naming competing products as anything other than "brand x."
Clinton is frankly being poorly served by campaign manager Robby Mook, who has worked on campaigns his entire career. Mook and his staff going up against master private-sector marketers like Trump and company have little chance to win in that arena. Heck, this group can't even put away Bernie Sanders, who, in contrast, is also a candidate who presents a lot more clear messaging and the appearance of a more authentic personality. And the worst news for Clinton is that it may already be too late to hire a new staff and present a new image to the public.
Of course, this election is still not over. But what is over is any notion that Trump is going to lose by some kind of landslide. I predict many more of my fellow pundits will come to this realization in the coming weeks.
Commentary by Jake Novak, supervising producer of "Power Lunch." Follow him on Twitter @jakejakeny.
For more insight from CNBC contributors, follow @CNBCopinion on Twitter.