3 innovators on reinvention, AI and metasearch's second act

They were once the rebels of online travel—the mavericks who built metasearch into a movement. On one stage at WiT Singapore’s The Next 20 conference, Gareth Williams (Skyscanner), Ross Veitch (Wego) and Kei Shibata (Travel.jp) reunited to look back on the early days of disruption and forward to what comes after metasearch.

Between them, they helped define the first generation of digital travel search—before artificial intelligence (AI), before super-apps, before Google rewrote the rules.

The trio were asked to respond to a statement from Aloke Bajpai of Ixigo, who said, “AI is the new meta. Meta is so screwed.” Is that true?

Origins: When travel was still a puzzle

For Williams, the idea that changed his life was almost a throwaway—“idea number 52.” Back in the early 2000s, three friends in Edinburgh, Scotland, were tossing around startup ideas. The spark came when Williams wrote a bit of code to find cheap flights for a ski holiday.

“Everyone says, ‘Don’t build a tool for yourself,’” he laughed. “But for us, that was the use case. We just wanted all the information in one place. It turned out a lot of other people did too.”

The early Skyscanner, like the best innovations, was born out of frustration and faith. “Unless you come from a travel insider’s point of view, you don’t know how fragmented the landscape is,” he said. “So, you build something you believe in.”

His design philosophy was simple: “Respect the airlines—they’re putting atoms in the air—but put the traveler first.”

The shape of metasearch: From marketplace to ownership

For Veitch, who founded Wego in Singapore in 2005, the model was always flexible. “We still run a vibrant marketplace,” he said, “But the fastest-growing part of our business today is the OTA side.”

He likened Wego’s evolution to Amazon or Alibaba—platforms that are both marketplaces and merchants. “Users should have a choice of who they book with,” Veitch said. “But if you can also serve them end-to-end—from search to booking to in-trip support—you add far more value.”

That hybrid approach, he believes, is essential in an AI-driven future. “If agentic systems are going to plan and manage trips for us, we need to be great at shopping and great at fulfillment.”

Williams agreed that travel’s big challenges haven’t changed, only the tools.

“Discovery, accuracy, support—those are still massive problems to solve,” he said. “But I think one thing we all missed was the social aspect. We built for solo users, even though travel is rarely a solo act.”

Kei Shibata: Betting on LINE, learning from loss

In Japan, Shibata’s journey was equally restless. His company, Travel.jp, has spent 25 years reinventing itself—from pure metasearch to business travel to the acquisition of Trip101 for a content play.

“The biggest risk was selling a big chunk of our shares to LINE, the messaging app. We made a full bet on LINE, we basically forgot the web. But it didn’t last long.”

He added, “That was a huge decision. But you have to take risks to stay relevant.”

Today, Shibata admits that “meta is almost broken.” He sees the current moment as a chance to rebuild.

“It’s time for big change. The dependency on Google is real—they’re like President Trump. You may not like him, but you can’t ignore him. If he doesn’t like you, you’re screwed.”

Many in the audience laughed, but his point was serious: Metasearch’s future depends on reducing dependency and reinventing value.

Asia rising: Lessons from the early days

When Williams first came to WiT in 2006, it was his first business trip to Asia. “It felt so exciting,” he recalled. “Europe was saturated, but Asia was full of fragmented markets—lots of airlines, languages and opportunities.”

Skyscanner’s obsession with coverage paid off. “Every time a new airline launched, we wanted every route,” he said. “Once you had that coverage, you automatically opened new markets.”

His vision on betting on Asia clearly paid off, with its $1.74 billion exit to Ctrip (now Trip.com Group) in 2016.

Veitch, meanwhile, learned the hard way that great products need great business models. “We built a beautiful product in Southeast Asia but not a great business. Low travel frequency, small baskets, few aggregators—it just didn’t scale.”

His solution was to pivot Wego’s focus to the Middle East, where “there were more players, more value to add.”

And Shibata? He turned his focus outward. “Japan was big for us, but we launched Trip101 overseas. The conditions that made meta succeed before—fragmentation, timing, no Google Travel—they’re gone now,” he said. “But in the AI era, those conditions might come back in a new form.”

AI: Metasearch’s second act

If metasearch was born out of information overload, AI may be its rebirth.

“Why did Google enter travel search?” Shibata asked. “Because they wanted to dominate intent. Now AI platforms—ChatGPT, Perplexity—want to do the same. They’re the new media. And like meta, they have to be neutral. So there’s a natural fit.”

Williams agreed that the old dream—the personal travel assistant—is finally within reach. “We talked about this 10 years ago,” he said. “Words are cheap, but now it’s possible. Every business traveler should have their own AI agent. The problem’s never been the tech—it’s that the customer is the finance department, not the traveler.”

What comes next: Passion, payments and purpose

When the talk turned to the future, all three founders looked outward again—at the next generation of entrepreneurs.

“I’m waiting for someone to solve B2B travel payments,” said Veitch. “Stablecoins will change this industry—take out the banks, remove friction. It’s a 12- to 18-month problem.”

Shibata’s passion lies in theme-based travel. “People travel now for very specific reasons—a concert, a snowboarding trip, a history obsession,” he said. He recently launched HistoricStays101, a site for travelers who want to “stay inside history—in a 200-year-old castle or ryokan.”

Williams, meanwhile, is coding again. After Skyscanner’s billion-dollar exit to Trip.com Group, he bought a plot of land, planted 40,000 trees in Scotland, took up motor racing (“I hate being a cliché ex-founder, but it’s fun”) and returned to his first love—computing.

“I’m learning maths again with ChatGPT,” he said. “I’m working on neuro-symbolic AI—and the idea of an intellectual firewall. We have tech firewalls; we need ones for thought. Otherwise our opinions will be shaped by big companies and governments.”

The founder’s journey: Charm, persistence and pain

Asked what they had to learn the hard way, the answers were revealing.

“I’m terrible at social media,” Shibata admitted. “But every entrepreneur needs charm—the smile, the charisma. Barry Diller, Masayoshi Son—they have it.”

Williams was more introspective. “I did it from a place of psychological maladjustment,” he said dryly. “I admire founders who do it from joy. I’d be too soft to do it again.”

Veitch nodded. “My happy place is still building product,” he said. “But I had to learn everything else—hiring, selling, pitching—all the people stuff.”

Heavy 'meta,' light hearts

In the final lightning round, Shibata called his journey “heavy but rock and roll.” Williams summed his up as “learning the value of teams.” Veitch? “Twisted.”

What would they tell their 27-year-old selves?

• Williams: “Don’t be overconfident.”
• Veitch: “Buy Bitcoin.”
• Shibata: “Be charming.”

And how do they see travel in 2045?

• Williams: “Leisure will dominate life.”
• Veitch: “It’ll be a much bigger part of everyone’s life.”
• Shibata: “The world’s biggest industry”

For three pioneers who built the early internet’s travel maps, it was clear they’re not done exploring.

This story originally appeared in WiT.