FutCrunch Makes 200 Thumbnails for One Video. At VidSummit 2026, He Explained Why

22 Steps to Grow from $500 to $10,000 on YouTube.pdf

Download Free Checklist

England's biggest football YouTuber walked onto the VidSummit 2026 stage and talked about what happens after the camera stops, mostly about thumbnails and what goes through his head when a video doesn't take off. "I think because I'm still delusional, I try and make everything work," he said.

We were in that room. We've dubbed FutCrunch's channel into 19 languages, after all. And he got really specific. For one video, he makes about 200 thumbnail concepts with AI, keeps 10 on the shortlist, and puts 3 into the test.

He also told the story of "the ugliest thumbnail in the history of thumbnails" and the million views it brought, explained why Shorts barely work for him, and remembered the plates he used to hand to his dad in a restaurant kitchen. We'll get to all of it.

What FutCrunch Talked About at VidSummit 2026, Four Years After His First Talk

VidSummit invited FutCrunch to share how he became England’s biggest football creator and how Shorts helped him reach 1 billion views during the World Cup. On Shorts, he was candid. He told the room they don't do much for him, because he doesn't see the people who watch them come over to his long videos.

That's worth checking on your own channel before you pour hours into Shorts. Pin a long video to each Short with YouTube's related video link, then look at that video's traffic sources. They'll tell you whether anyone comes over.

Derral Eves had his own reason to bring up thumbnails. Four years earlier, at VidSummit 2022, FutCrunch gave a talk called "How To Go From 1000 Views A Month To 38 Million", and Derral made his son watch it. Derral still remembers one example from that talk. An older video was getting views from suggested, so FutCrunch gave it a new thumbnail. Derral called it "the ugliest thumbnail in the history of thumbnails," and it brought over a million extra views.

So the chat turned to packaging, and from there to titles in 19 languages, a $10,000 trip to Paris and his dad. The 19 languages come from our own interview with FutCrunch's COO, not from the stage.

How FutCrunch Makes Thumbnails Now: 200 AI Concepts, Then a Designer

Six months before VidSummit, FutCrunch was still drawing his thumbnail ideas with the brush tool in Photoshop. "It was horrible," he admitted. The sketches went to friends and the team, three got picked, and a designer made them properly.

Then one day the designer couldn't get a concept right. "Bro, let me use AI," FutCrunch said. The thumbnail the designer finished came out almost the same as the AI version, and he asked himself, "So what is my paint doing for?" The next three hours, the time that used to go on sketching, gave him eight AI thumbnails.

Now it's closer to 200 per video, and every video starts the same way.

Before the first prompt, he types the video's topic into YouTube search and looks at how other creators packaged it, so the AI gets a reference. Then come the angles, the ten he likes best, and the thumbnail editors who finish them. The AI still gets things wrong. "Every now and then it doesn't look like me," he said about his own face.

His face shows up less these days anyway. When Haaland is in the video, Haaland goes on the thumbnail, and FutCrunch keeps versions with himself and without for the test to decide.

How FutCrunch A/B Tests Thumbnails: Four Hours, Then Let the Video Breathe

Five years ago, YouTube had no A/B test for thumbnails. So when a video FutCrunch rated 10 out of 10 flatlined, he did the only thing he could. He made a new thumbnail, swapped it in, and just sat there staring at the real-time numbers. "Maybe I'm toxic. I don't know," he joked. The new one took the video "from like a flatline video to a whole other level."

Now the test is built in. Every video goes up with at least three thumbnails he's happy with, and about four hours after release he usually knows which one leads. The leader then goes up against the rest of his shortlist, and after that he leaves it alone. "One thing that's underrated is letting the video breathe."

Three rounds, then he lets it go

He still checks real-time numbers every day. If a video isn't in his top four out of ten in YouTube Studio's ranking, a whole day goes on ten new concepts. Three rounds like that, and he stops.

A month or two later, when he has free time, he comes back and tries again.

The winner can change, too. FutCrunch has rerun a test a month later and watched a different thumbnail come out on top. Derral thinks it's because after 28 to 30 days, the video reaches new people. Your most engaged fans watch first. Later, YouTube shows the video to people who may not know you, and they react to something else.

Derral's way: test it on an old video first

Derral doesn't split-test at release. He goes out with his strongest title and thumbnail, keeps two backups, and swaps only if the video falls below his usual numbers. A new thumbnail idea gets a trial run on an older video about the same topic first. If it works there, he switches the old video back and saves the thumbnail for the new release. "You already know if it's going to work with the traffic type and the device type."

If you use YouTube's own test, don't expect an answer in four hours. Test & Compare can take a few days or up to two weeks, and it picks the version with the most watch time (YouTube Help). FutCrunch reads his four-hour signal from real-time views on a 24-million-subscriber channel. A smaller channel needs the full test. We compared AI tools for thumbnail A/B tests if you want more than the built-in one.

Why FutCrunch Changes Thumbnails but Rarely Titles

Then Derral asked whether he ever changes the title. Almost never. A thumbnail takes up most of the space a video gets on YouTube, so that's where the wins are. And he tries to get the title right before the video goes out.

The other reason is a lot of extra work. "Changing the title in English, I just have to change it in every single other language," he said.

We know that part well. AIR Media-Tech launched FutCrunch's multi-audio tracks in 19 languages, from Spanish and German to Japanese, Arabic and Thai. The team added them a few at a time, and each batch had to pay for itself. "Make sure it's giving back the money we spent on dubbing," COO Moises Valor told us in our interview about the channel's 19 languages.

Derral pushed further. Since August 2026, channels with advanced features can upload a separate thumbnail for each audio language, and FutCrunch's audience is worldwide. Why not use it? FutCrunch did the math out loud. AI still can't get the small details right in every language, designers charge a lot, and he may test 20 thumbnails on a video without a clear winner.

He'd rather keep his team small, so everyone on it earns more.

For a channel just starting with dubs, we'd put it differently. A thumbnail with English text over a Spanish track tells a Spanish-speaking viewer the video isn't for them. If per-language versions are too much, a thumbnail with no text works in every language. And settle the title before it goes to translation, because after that every fix costs one change per language.

FutCrunch is waiting for the next step. At Made on YouTube in September, YouTube announced dynamic thumbnails. You upload three options, and YouTube shows each audience segment the one that suits it best. "I am looking forward to that," he said.

The 67-Million-View Haaland Video Began as Someone Else's Idea

When FutCrunch heard he might get to meet Erling Haaland, he asked himself, "If I were to get only one second with this dude, what would I do?" The answer was already on YouTube. Another creator had surprised LeBron James with custom shoes, and the video did well.

FutCrunch made it bigger. He flew to Japan, brought custom shoes for Haaland and gifts for his teammates, and turned the handover into a chase that only ended when Haaland had the shoes. It became the most expensive video he's made, and a risk, because nobody had done it before.

"I Surprised Haaland With Custom Shoes" now has about 67 million views. OutlierKit puts it at the top of the channel, about 7.5 times its usual result.

That's the move we'd steal. Find a moment that already worked on someone else's channel, then ask what would make yours bigger, like a longer journey, more people in it or a clear finish line.

From FIFA to Real Life: Ten Times the Views, Ten Times the Work

FutCrunch also runs a FIFA channel. One person and three editors make it, and it earns about as much as his main channel, where some ten people travel the world for real-life football videos. "Which is insane," he said. "To this point, still baffles me."

So why leave the easy money? Because he felt stuck. He had become the biggest FIFA creator in the world, and making the same videos would burn him out. Other channels had already gone from FIFA gameplay to real-life football videos, so he made the same switch, step by step.

First came FIFA videos with a real-life twist. He stuck player cards on a real goal, shot at them, and bought whoever he hit for his FIFA squad. Those got about ten times the views of his gaming videos and took about ten times as long to make. Then real-life videos started bringing a million views on the first day and two million in a week, while gaming ones got 100,000 a day. He moved over completely.

The first big one began as an excuse. A Real Madrid fan, he wanted to see the 2022 Champions League final in Paris, so he asked a friend how to make it a business expense. Tickets were about $5,000 each. They spent $10,000, went to Paris, and filmed with no script, piecing the story together over the week. Editing took longer than shooting, and FutCrunch cut the first half himself. He says it became the most-viewed Champions League final video on YouTube.

If you're stuck in one format, that's the order we'd copy. Start with a hybrid. Keep what your audience knows and add one real-life element. Then compare views and hours per video before you switch.

"Find Your Why": The Restaurant Story He'd Told Very Few People

Derral saved the big question for last and asked what FutCrunch would tell everyone in the room, whatever stage they're at. Know why you're doing this, he said. Without a reason, three years in, you'll wonder why you're putting in all the time, while someone else wakes up "with a fire that you don't have."

Then he told the room something he said he'd told very few people. Years ago, FutCrunch, whose real name is Santiago Vázquez, worked as a waiter in a restaurant in England. His dad had been a lawyer in Venezuela, but the degree meant nothing in England, so he went back to university and took a job washing dishes. In the same restaurant.

That went on for two years. Every day, he was fighting for his dad and his mom, and anyone competing with him "had no chance." He made videos for three years while still working that job.

When he could finally retire his parents, the reason was gone, and he had to find a new one. Now it's to make the most of his career for as long as he can. Being the biggest, he said, ends up being just a number.

We'd add only one thing. Write your reason down in one sentence, and when it's done, the way his was once his parents retired, write the next one.

What to Take From FutCrunch's Session for Your Channel

FutCrunch has a team, a design budget, and 24 million subscribers. Most channels have none of that. Here's what we'd take from his session for a channel of any size.

1. Three real options per video. Make them different ideas, so the test tells you something, then leave the winner alone for a few weeks.
2. A second thumbnail for old videos. A month or two later, a video meets new people. When we audited My Untold Story, a faceless storytelling channel, we took the horror-bait faces and red arrows off dozens of its thumbnails and titles, and CTR went from 5.6% to 6.4%.
3. Shorts that earn their place. Link each one to a long video and see if anyone comes over.
4. The title first, then the dub. Translate a title you won't touch again, test the thumbnail, and localize it only for languages that already pay back.