Let's say you have to buy 400 carts-worth of groceries from the store. It doesn't matter why; maybe you're den mother for the world's neediest group of Cub Scouts. Anyway, you have 400 carts to buy, your cart holds one cart's worth -- and the closest store, a half hour away, is only open for 12 hours, one day a year.

This isn't a math question that you're required to solve, although you're welcome to. It's an illustration. The task, as presented, is impossible.

Let's say, though, that we open grocery stores closer to your house for a few hours every week. Then, on top of that, you can order groceries online. You will probably still end up running a lot of trips to the main store on the day that it's open, but you can get a lot done beforehand.

This is an analogy.

In 1996, nearly 90 percent of voters in the November elections voted by trudging to their polling places on Election Day, according to the Census Bureau. In 2014, over 30 percent voted early or voted by absentee ballot. In 2012, nearly a third of votes in the general election were not cast at an actual polling place. Lots of people got their shopping done early.

But the analogy here isn't the voters. It's not the voters who need to go to the polls 400 times (but I look forward to your funny Chicago/ACORN jokes). It's political campaigns. And they, like our humble den mother, would rather get as many votes in before the fact as they can, too.

A vote's a vote. It doesn't matter when it is cast. So given the opportunity to do turnout for a month before Election Day, versus just on the day itself, campaigns take advantage.

By the time you read this, for example, I can definitively assert that more than 1.1 million votes will be cast in the Republican primary in Florida. I know this because Daniel Smith, a.k.a. Election Smith, knows this. Smith's analysis of the early vote tells us a lot about what we can expect in the state on Tuesday. One example: Only about 59,000 of those 1.1 million voters went to the polls in each federal election since 2008. Which suggests a lot of less regular voters. Which suggests a decent number of Donald Trump voters.

There's other evidence to that effect. We noted two weeks ago that Marco Rubio was probably already losing in the state at that point, because he was trailing Trump in the polls as early voting began. A poll from Monmouth University last week suggested that, with a fifth of Republicans surveyed saying they're already voted, Rubio actually lead in the early vote. In a new poll, a third of the electorate reports having voted (suggesting a voter universe of nearly 3.5 million people) -- and Trump now has the lead.

Those votes are groceries in the bank. If Donald Trump were to defect to Libya on Tuesday morning and no one came out to vote for him at the polls, he'd still have a huge chunk of support banked. In Georgia, for example, Trump's early-vote margins were enough for him to beat John Kasich and Ben Carson even if no one had voted for him on primary day.

The odds are good that Trump's benefiting more from his strong position in the polls than from a robust organizing effort of the sort that he, so far, hasn't demonstrated. So how do strategic campaigns get people to vote early, allowing them to put votes from supporters in the bank? (This is a term of art: "banking votes.")

Let's look at Nebraska.

In Nebraska's Democratic caucuses earlier this month, Bernie Sanders beat Hillary Clinton by 4,780 votes, earning 57.1 percent of the votes cast and 15 delegates to Clinton's 10, according to U.S. Election Atlas. Unlike most caucuses, Nebraska allowed people to vote by absentee this year. And at least 2,900 Democrats voted for Clinton by absentee vote, according to the campaign -- enough that, had they not done so, Sanders would likely have come away with a 7-delegate margin, not a 5-delegate one. Those are two delegates that, somewhere down the line, might end up being essential.

How does the Clinton campaign know that 2,900 people voted for them absentee? Because they sent voters in a certain target group (or "universe") a letter from former senator Bob Kerrey (D-Neb.) asking them to fill out an enclosed form and return it to the state party in the enclosed envelope. Which is where the 2,900 figure comes from: That's how many envelopes were returned.

More than those 2,900 voted for Clinton by absentee, it's safe to assume. More than 6,200 absentee votes were cast according to Iowa Starting Line, a blog about primary politics. But the campaign worked to get those 2,900 votes in. It wasn't just the letters.

"We mailed folks, we called them, we knocked on their doors, etc., to chase the mailing we did," said Marlon Marshall, Clinton's director of state campaigns and political engagement when we spoke by phone last week. So that's a piece of mail, a phone call and a knock on the door (as much as possible) to encourage Nebraska Democrats to vote before the day of the caucus. "We were able to then focus our field program" -- the campaign's get-out-the-vote effort -- "on those who did not vote absentee," Marshall said, "which shrinks the universe of those that you want to talk to and doors you're knocking on in those last few days." Going grocery shopping early.

Clinton's team has been doing something similar in Florida -- and elsewhere. "We look at every single state to see how we can make sure that we'll run a program that will maximize participation in general," Marshall said. In Florida, Smith estimates that 820,000 Democrats have already voted. A slightly higher percentage of them voted in all four of the most recent general elections than on the GOP side.

Campaigns spend a lot on turnout, so tightly tailored efforts to bank votes from targeted voters early can pay dividends. In January, the last month before Iowa, the four remaining candidates on the Republican side spent $2.1 million on staff, phone lines, printing and rent, by our calculations -- four of the main cost areas for voter turnout efforts. Ted Cruz spent $68,000 that month alone on identifying how specific voters were leaning, in order to figure out who to target for turnout efforts. His efforts paid off in Iowa.

But not all of Cruz's money was necessarily invested in Iowa or New Hampshire. Plenty of shopping that needed to be done in early voting states, too.