In Silicon Valley, there's a well-known blueprint for a startup. It should start with a simple concept. Then, it should do a lot of "A/B testing" and "iterating," which is fancy-speak for throwing a lot of stuff against the wall and seeing what sticks. Hopefully, with enough iterating, the startup will achieve what's known as "product-market fit" — the product has evolved enough that now people really like it. That's when you're supposed to step on the gas: Now that people like your product, you should ramp up marketing and production.

Of course, the goal is to "disrupt" the competition. While in business scholarship the word has a very specific definition that's not worth getting into here, the word has become a shorthand for producing something so new and out of left field that it discombobulates the competition and allows you to grab the market for yourself in short order.

All of which is a way of saying Donald Trump is the political equivalent of a Silicon Valley startup. Not because his campaign is particularly high-tech — it's the opposite. But because its strategy looks like the above playbook.

A lot has been said about Trump's incoherent speech and, more broadly, the seeming lack of a communications or messaging strategy. Typically, campaigns come up with their message and strategy early on and roll it out. Of course they adjust and tweak on the fly, but they rarely make big changes — or else it's a sure sign of trouble. Marco Rubio had his optimistic message about a new American century and opportunity, and basically never deviated from it and stayed on message. Only late in the campaign did he pivot to attacking Donald Trump, and even then, while those sections were emphasized in cable news, his basic stump speech remained unchanged besides the Trump stuff.

Trump's incoherence makes a lot more sense if you think of it as A/B testing. Trump uses his rally, his Twitter accounts, and his interviews to try things out and see what works. Think of the way he would cycle between demeaning nicknames for his opponents until one stuck.

Of course, this is a perfect strategy for the social media and cable-TV driven news cycle, when the premium is on novelty and outrageousness. This week Trump has been parroting the conspiracy theory that the Clintons murdered their aide Vince Foster. It seems strange given how many actual scandals those two have had. But he's just trying things out.

Once you find something that sticks, you have to double down. Trump's "idea" for a Muslim ban was originally a botched non-answer to a question. But when the outrage machine started spinning, Trump doubled down, airing ads in favor of the idea. Even now, despite his supposed general election "pivot," he's still playing it up.

A Silicon Valley startup is supposed to be lean, and that's certainly what the Trump campaign has been, spending a lot less money than other campaigns. Why should you when you have a system designed to dominate the news cycle? Startups find unusual routes to the market that allow them to do a lot more with stretched resources.

And it was certainly disruptive. None of the other candidates knew what hit them.

What does it mean for the general election?

Well, first, disruptive startups usually do a lot better than anybody expects. Facebook was a laughingstock all the way to a billion users and multi-billion dollar profits. It's still puzzling for anyone older than 15 why anyone would use Snapchat. Trump's strategy is a perfect fit for a post-truth candidate in a post-truth age. Hillary Clinton has a message that she has to stick to. Trump can say whatever will get him the headlines.

Second, don't underestimate Trump, and especially his way to turn weaknesses into strengths. His ramblingness and lack of focus seem like a weakness, but they're actually a ruthlessly effective strategy.