There are some elements of a modern campaign that ought to be questioned. I’m thinking in particular of the efficacy of television advertisements and the way political operatives make a killing on commissions for buying millions of dollars worth of 30-second spots. This is an expensive way of reaching voters, and it’s very inefficient not just because of the cost, but because it’s a blunt instrument. In the modern world, there are ways to get your message targeted to just the eyeballs you want to reach rather than flooding your message to as many eyeballs as possible. How many people even sit through television advertisements these days?

But, there are other age-old campaigning tactics and strategies that are just as important today as they ever were. One is identifying your supporters. Another is getting your supporters registered to vote. Another is getting as many registered supporters as you can to actually cast a ballot. All of these work better if you have a good voter contact organization. They work better if you can collect, preserve, and disseminate information. This is what makes up a good ground game, and a lot goes into it. Modern campaigns do a lot of modeling of the electorate, much of which goes far beyond simply looking at voter lists and prior voting behavior. If you know that someone owns a gun or drives a Suburu, that’s helps you figure out if it’s worth sending someone to their door to talk about the campaign.

And you want to know what people said when your people knocked on their door. You want to know if they were home, if they committed to voting for your candidate, or they instead set their Rottweilers loose on your organizers.

These kind of data flows are maintained by sophisticated software and apps, as it’s just as important to collect the information from the field as it is to turn around and send it back to the field so organizers can have an efficient and pleasant experience.

How important is all this stuff to determining the outcome of a presidential election? Sometimes, it’s not really important at all. There’s nothing a better field organization could have done to save Walter Mondale or Bob Dole’s campaigns. The primary responsibility for winning lies with the candidate, and there are economic, ideological, and cyclical factors that can sometimes overwhelm the most dedicated work of a campaign and its organizers and volunteers. But, if one campaign has a substantial field advantage, it can be decisive in a close election.

So, unless Trump is going to win this election in a walk based on winning the argument against Hillary Clinton, he’s going to want to be competitive in the field. And he is not going to be competitive in the field because he really has no campaign at all.

Even at the higher level of media communications, Trump has a skeleton crew.

In reporting on Trump’s operation, NBC News talked to three Trump aides and two sources working closely alongside the campaign, all of whom requested anonymity in order speak freely.

Veteran operatives are shocked by the campaign’s failure to fill key roles. There is no communications team to deal with the hundreds of media outlets covering the race, no rapid response director to quickly rebut attacks and launch new ones, and a limited cast of surrogates who lack a cohesive message.

“They don’t or can’t cover it all, and there are things that happen that need to be addressed immediately and don’t get addressed at all, and that hurts the candidate,” a source within the campaign groused last month.

The campaign is bringing on a new senior staffer Jim Murphy, as first reported by The New York Times, and a source said more communications hires are expected to follow. But they lag far behind the Clinton campaign, which has over a dozen senior staff dedicated to communications as well as teams devoted to modern data and analytics, an area where Trump is publicly skeptical of hiring. In addition, Clinton enjoys support from established super PACs like Correct The Record and American Bridge that respond to attacks and promote opposition research.

Perhaps Trump can make up for a deficit in the communications battle by staffing up and by utilizing his unique ability to drive news coverage. But he can’t overcome the lack of any real field operation. If the battle continues to be this lopsided on the ground, it’ll cost Trump three-to-five percentage points in the key battleground states, and that’s not something a Republican candidate is likely to survive if the election is close at all. The way the Electoral College is set up, the best a Republican can normally hope to do is win by the narrowest of margins. If Gore had been awarded Florida in 2000, he would have been president, and Kerry would have been president if he had won Ohio, despite losing the popular vote. The 2004 election is the high-water mark for a Republican campaign since 1988, and demographics are less favorable today than they were then. And, in 2004, Bush won by the skin of his teeth by producing a superior ground game in Ohio (and, by having a very cooperative state government working to diminish Democratic turnout).

I don’t think Trump can muscle home that kind of narrow victory. His only hope is to overwhelm Clinton with a message that people buy without any prompting from organizers. So, he better figure out how to fix his practically non-existent communications shop. That’s his only hope.