Okay, I didn’t do so well with my 2020 predictions. I was wrong that fact-checkers would win a Nobel Prize. And no, Brad Pitt and Charlize Theron didn’t star in a movie about fact-checking, nor did Taylor Swift make it the focus of a hit song that she performed at the Super Bowl.

But there was a lot of fact-checking in 2020! Indeed, there was an avalanche of Pinocchios and Pants on Fire ratings, plus tremendous growth in embedded fact-checks, the practice when reporters assess claims right in their news stories with words such as “baseless” or “unfounded.” And during the presidential debates, a few outlets (including PolitiFact, the website I started) had promising results with experiments in live on-screen fact-checking.

Still, I’m going to be more realistic with my 2021 predictions. I think the future of fact-check journalism is all about structured data.

Sound a little dull? It’s actually an idea that has been around for a while. It goes back to 2006, when a visionary journalist-developer named Adrian Holovaty wrote an essay titled “A fundamental way newspaper sites need to change,” which made the case that journalism would be more valuable presented in structured form, like data. The essay caught the imagination of Matt Waite, my collaborator at PolitiFact, and it inspired us both.

The journalism-as-structured-data revolution succeeded in a few places, like PolitiFact and Chris and Laura Amico’s Homicide Watch, but it hasn’t succeeded on a broad scale. Journalists are storytellers accustomed to an old story form, and they’ve had trouble adapting their work to a structured approach.

But suddenly the time is right for structured journalism, because our chaotic battle over misinformation is a perfect opportunity to take advantage of fact-checking as data. Liars say the same things over and over, which makes the fact-check we wrote last week or last month valuable for an extended period. So if fact-checkers add some simple tags to index their articles, search engines and other platforms can match the lie with the correction.

Five years ago, my team at the Duke Reporters’ Lab worked with Google, Jigsaw, and Schema.org to create just such a product, a tagging system we called ClaimReview. Most fact-checkers around the world now add ClaimReview tags to their articles and then Google and YouTube and Facebook — and anyone — can find those 70,000 fact-checks through an open database.

I think of it as the hidden plumbing of fact-checking that makes it easier to get facts about falsehoods.

ClaimReview was big in 2020. YouTube used it to highlight fact-checks in its search results. Google, which uses ClaimReview for search results and Google News, said users saw more than 4 billion fact-checks in its products in the first eight months of the year. Bing also uses ClaimReview, we use it to power our experimental live video app Squash, and it could be a big help to Twitter.

Next year, we’ll be testing a new kind of tagging system for fact-checks of fake videos and images. Our Duke team has been working with fact-checkers and the tech platforms to develop a sibling of ClaimReview that we call MediaReview, which creates a common language to describe deepfakes and other bogus videos and images.

By consistently using terms such as “missing context” and “edited,” the fact-checkers can provide Google, Facebook, Twitter, or any other platform with instant information about what’s false or misleading about a video or image. The platforms can then make quick decisions about what to do with that content — they can reduce its spread, delete it, or leave it alone.

MediaReview isn’t sexy; plumbing rarely is. It’s just structure that will help solve a complex problem. But it will be big in 2021.

And then in 2022, Taylor Swift can write a song about it.