Both of them were up at the wrong hour. That part came first.

Ohio, a little after one in the morning. Jake reheats a coffee he won't drink and thumbs the same four apps in a circle, the way you open a fridge you already know is empty.

Lisbon, just after six. Maria is awake before the city is, again, watching the river go from black to grey. She has always been early. Her whole life people have told her to relax, that the right person turns up when you stop looking. She stopped looking years ago. Nobody turned up.

Two people, both awake when they shouldn't be, in the one century where that finally meant something.

the first few weeks

They did not fall in love. Not yet. They just kept turning up at the seam between their days.

He started going to bed later. She started waking up earlier. Somewhere in the middle was a ninety-minute window, late for him and barely morning for her, and without ever agreeing to it they began to treat that window as a place. A room with a door. Theirs.

This is the unglamorous truth about the most romantic thing that ever happened to either of them: it was mostly two tired people, texting.

Jake and Maria had exactly one thing in common: an app, and the wrong hour.

A generation ago that was a disqualification. You don't even know each other. Now it is just how the story starts, for almost forty in a hundred of us.

eight months of good morning, good night

By then everyone had a verdict.

And the worst part: they had a point. Jake and Maria knew each other the way you know a voice in the dark. Completely, and not at all.

This is the oldest job in the world talking. The aunt, the priest, the friend who knew a guy, every gatekeeper who ever asked but who are his people? For ten thousand years that question kept us safe, and kept us apart. Jake was about to do the one thing it was built to prevent.

the numbers behind Jake and Maria

Jake and Maria are drawn, but the numbers around them come from the How Couples Meet and Stay Together survey, which the Stanford sociologist Michael Rosenfeld and his colleagues published in PNAS in 2019, from Pew Research Center, and from a 2017 paper by the economists Josué Ortega and Philipp Hergovich called The Strength of Absent Ties.

In that survey, 2 percent of straight couples who got together in 1995 had met online, a share that climbed to 22 percent by 2009 and 39 percent by 2017, while the share who met through friends fell from 33 percent to 18 percent over the same years.

The two lines crossed in 2013, at 30 percent online against 20 percent through friends, and by 2024 The Knot's survey of couples put the online share at 60 percent, while for same-sex couples meeting online was already the most common route in 2009, at 61 percent, and reached 65 percent by 2017.

Of the couples who met online, 89 percent had no friend in common before they met, according to the 2017 wave of the same survey, which is the whole point of this story, because no cousin or colleague could have put a man in Ohio in touch with a woman in Lisbon.

The share of American newlyweds marrying someone of a different race or ethnicity rose from 3 percent in 1967 to 19 percent in 2019 in the Pew figures this page uses, and when Ortega and Hergovich modelled a society in which strangers start meeting through apps, their model predicted a rise in interracial marriage of exactly this kind.