Even prior to clinching the Libertarian Party’s nomination for president on Sunday, former Republican Gov. Gary Johnson had been included in three major polls as the likely alternative to Donald Trump and Hillary Clinton, which consistently indicated that his level of support is hovering around 10% — an impressive figure for any third-party candidate. Now that Johnson is the nominee, and succeeded in persuading the party’s delegates to nominate his pick for a running mate, former Republican Gov. William Weld, his poll numbers can be expected to be the subject of intense scrutiny and discussion.

But what do the numbers really mean, from a strategic perspective?

The earliest poll of interest, conducted by Monmouth University in March, included 848 registered voters, and has a margin of error of 3.4%. 34% chose Trump, 42% Clinton, 11% Johnson, 1% “other,” 7% “no one,” and 5% weren’t sure. Proving that the Monmouth poll wasn’t a fluke, a Fox News poll of registered voters, released on May 18, found that 42% supported Trump, 39% Clinton, 10% Johnson, and 2% “other,” while 5% said they wouldn’t vote, and 3% didn’t know. The margin of error was 3%. In yet another poll of registered voters, this one published by Morning Consult, Trump received 35%, Clinton 38%, and Johnson 10%, with 17% undecided.

Interestingly, although the Fox poll seems a bit skewed towards Trump (which might not surprise critics of the conservative-leaning network), Johnson’s position around 10% seems consistent and presumably accurate, and has remained stable over the course of several weeks.

Johnson and his supporters are rightly excited; the figures are about double those from the 2012 campaign season, when he was also the LP nominee, and they view 10% as a starting point to build on during the next six months. The target of 15% — a necessary step towards getting into the fall presidential debates, and a crucial objective in any third-party candidate’s campaign strategy – certainly seems to be within reach.

However, it would be a mistake to assume that one in ten registered voters knows who Johnson is, agrees with his platform, and has decided to vote for him in November. Considering the well-documented unpopularity of both Trump and Clinton even among their supporters, it’s more likely that 10% of voters are simply willing to vote for just about anyone other than the Republican and Democratic options. Or they’re at least willing to say they are, which is, of course, quite different than actually doing it.

A prime example of this polling phenomenon went viral last summer, when the independent joke candidate Deez Nuts received 9% in a North Carolina poll of likely voters. Such protest votes, as pollsters describe them, are rarely duplicated in election results.

Still, Johnson isn’t Deez Nuts (though some dissatisfied Libertarians might choose the latter if given the chance). He’s a two-term Republican governor from a blue state who initially won election as a political outsider, and his level of support is likely to improve amidst a wave of media attention and an expected influx of campaign contributions. But the most significant obstacle he faces is one that he can do little about: Pollsters choose which candidates to include, and most are choosing Trump and Clinton.

In addition to leading voters to believe that they only have two options, which naturally influences their opinions, two-way polls (including those that offer “other” and “undecided” as options) perpetuate a vicious cycle: If third-party candidates are excluded from polls, they can’t poll at 15%; if they don’t poll at 15%, they’re excluded from debates; if they’re excluded from debates, they don’t do well in elections; and, of course, if they don’t do well in elections, pollsters don’t include them in polls. And so on.

Assuming that Johnson’s strategy is geared towards breaking this cycle and getting onto a debate stage with Trump and Clinton, his executive experience and moderate brand of libertarianism, though considered weaknesses by many younger Libertarians, may be his greatest strengths, for they increase his chances of being included in mainstream polls which might not have included a lesser-known candidate, Libertarian nominee or not.

To convince both the media and the general public that he’s more than a Deez Nuts – more than an easy way for frustrated voters to vent about the Republican and Democratic choices before grudgingly backing one of them or staying home on Election Day – Johnson will need to do a bit better than 10% in the next polls that include him. Not a lot better; even two or three percentage points would demonstrate growth, trigger another round of news stories and interviews, and build name recognition.

Since voters who have already decided to support Trump or Clinton are unlikely to change their minds (some will, but Johnson probably won’t pick off enough to make a major difference, barring extraordinary circumstances), he must focus on a specific group of voters: Undecided millennials. Not only are younger voters more likely to support an independent candidate – a recent Data Targeting poll found that an unprecedented 91% of voters under 29 favor having an independent candidate on the presidential ballot this year – they’re also more likely to be undecided. Many are still fiercely devoted to Sen. Bernie Sanders, but following Clinton’s nomination, they’ll be more open to an alternative than any other group of voters. Though the Vermont socialist might seem like an unlikely ally for Libertarians, Johnson has already emphasized their similarities when it comes to foreign policy and some social policies.

If Johnson can win support among undecided and disillusioned millennials, the next big obstacle between him and higher poll numbers might be a bit tougher to overcome: Their tendency to ignore calls from unfamiliar numbers, including pollsters.

Perhaps an #AnswerYourPhone social media campaign will be the next big thing in Libertarian circles.