What 25 years in travel tech taught me about the future of personalization
Twenty-five years ago, my friend, Emmanuel Ciolfi, and I were just two broke dreamers sharing a cramped London flat and a single laptop. What we lacked in resources, we made up for in vision. We launched London Theatre Direct, the UK's first online theatre ticketing platform. The internet was still the Wild West back then; Google was barely a blip, Facebook didn't exist, and smartphones were years away. But I had this gut feeling that technology was about to turn the travel and entertainment world upside down.
Fast forward to 2025 and I'm still amazed at how far we've come. This isn't just evolution—it's a revolution. We've entered the age of agentic AI, and it's changing everything.
BREAKING FREE FROM THE ONE-SIZE-FITS-ALL TRAVEL BOX
Traditional online travel agencies were once the disruptors, but in my opinion, they've always treated travelers like interchangeable data points rather than actual humans with quirky preferences and spontaneous desires.
Today's travelers (myself included) want something that actually feels intuitive. Not just systems that know we've been to Paris three times, but ones that understand why we keep going back. Maybe it's that specific patisserie on the corner of Rue Cler that makes those perfect pain au chocolat we can't stop thinking about.
This is where agentic AI is changing the game.
When we built tickadoo, I wanted technology that didn't just passively recommend options from a static menu but actively anticipated what might delight users. I still get goosebumps watching users' faces light up when they realize our platform isn't just responding to what they've explicitly asked for—it's also reading between the lines of their past choices to suggest experiences they didn't even know they wanted.
THIS ISN'T SOME FAR-OFF DREAM—IT'S HAPPENING NOW
What moves me is the democratization happening here. The kind of personalized service that once required a platinum credit card or knowing someone who knows someone is becoming available to everyone. That's the mission that gets me out of bed every morning—making the exceptional expected and building loyalty through genuine understanding rather than marketing tricks.
I've had countless conversations with executives at legacy travel companies, and I see the fear in their eyes. Embracing agentic AI isn't just about updating some software—it requires completely rethinking how they relate to customers.
The companies brave enough to make this leap won't just survive—they'll lead. Those who cling to outdated models? Well, I've seen enough disruption in my 25 years to know how that story ends.
BEYOND BOARDING PASSES AND HOTEL KEYS
This transformation extends far beyond my industry. Health care systems are starting to predict and prevent issues before they arise. Entertainment platforms are beginning to understand not just what genres consumers like, but what emotional experiences they're craving on a rainy Tuesday evening.
At tickadoo, our AI Concierge doesn't just mechanically suggest activities. It learns the subtleties of our users' preferences. It notices that while they enjoy museums, they specifically light up at interactive exhibits. It remembers that they prefer evening shows after a relaxed dinner, not rushed pre-theatre meals. These aren't just conveniences—they fundamentally change our relationship with technology.
HOW TO GET STARTED WITH AGENTIC AI
Agentic AI isn't just another tech upgrade—it's a fundamental shift in how businesses connect with customers. Here are practical steps to help you begin this transformation:
1. Start with vision, not technology. Define what anticipatory service means in your industry. Is it predicting health care risks before symptoms appear? Is it understanding emotional states for entertainment recommendations? Is it anticipating style preferences in retail?
2. Invest in deeper customer understanding. Move beyond transactions to behaviors and emotions. Tools like sentiment analysis and journey mapping reveal the subtle preferences that AI can transform into compelling differentiation.
3. Launch focused pilot projects. Retailers can test recommendations based on lifestyle signals rather than just purchase history. Health care providers might flag subtle health changes proactively. Entertainment brands can suggest content based on real-time emotional engagement.
4. Blend human expertise with AI insights. The magic happens when empowered humans use AI-driven intelligence to create extraordinary experiences. Train your teams to interpret and apply these insights with empathy.
5. Create a culture of experimentation. The biggest barrier isn't technical but cultural. Value insights gained through testing as much as immediate outcomes. Build an environment where rapid learning is celebrated.
6. Prioritize transparency. As AI becomes more integrated, customers want to understand how it works. Be clear about how you use AI and how it respects privacy. Transparency transforms skepticism into trust.
7. Find strategic partners. You don't need to build everything in-house. Collaborate with startups and specialists to accelerate learning, reduce investment risk, and shorten time-to-value.
8. Develop new metrics. Traditional KPIs may miss the real value of agentic AI. Consider measuring predictive accuracy, emotional satisfaction, or proactive retention alongside conventional metrics.
Sometimes when I'm jet-lagged in yet another hotel room, reviewing metrics on my tablet, I think back to that cramped London flat where it all started. That scrappy operation with a shared laptop feels worlds away from our current AI infrastructure. But the core excitement, that belief that technology can make travel and entertainment more accessible, more personal, more human, remains exactly the same.
With our recent launch of memberships, we're doubling down on personalization. This isn't just the future I predicted all those years ago. It's the one I've been working toward, step by step, for a quarter century. And now that it's here? I can't wait to see where we go next.
Try Our AI Features
Explore what Daily8 AI can do for you:
Comments
No comments yet...
Related Articles
Forbes
24 minutes ago
- Forbes
5 Times Conventional Startup Advice Backfires
In this article we give famous real-world examples of successfully breaking conventional startup ... More advice. Sometimes, the widely shared mantras in the startup world passed down from investors, accelerators, or successful founders can turn into liabilities when applied too broadly or in the wrong context. This article examines five pieces of startup advice that are commonly repeated and shows how they can backfire when misunderstood or taken too far. 1. 'Move Fast And Break Things' Can Break The Wrong Things Originally coined by Facebook, this phrase became shorthand for prioritizing speed over perfection. And in many early-stage contexts, it makes sense - speed allows you to iterate, test assumptions, and beat slower-moving competitors. This principle is engraved in the startup DNA, especially in the early startup stages when you need to make mistakes fast so that you learn fast and find the right direction of iteration, hopefully leading you to product-market fit before your competitors. But when applied without discipline, 'move fast' can result in poor technical foundations, half-baked product decisions, and organizational debt that slows down future progress. In critical systems—healthtech, fintech, or any sector involving real-world risk—moving fast can be dangerous. Even in consumer SaaS, shipping without user research or QA can alienate early adopters. Many early-stage fintech companies learned this the hard way by skipping compliance considerations in their early days, only to face regulatory scrutiny later. Stripe's early team, on the other hand, built a developer-friendly API and took care of banking compliance from day one. The better version of this mantra is to move fast on learning, not just building. Speed should be tied to clarity of assumptions and feedback loops, not chaos. 2. 'Fail Fast' Can Lead To Premature Abandonment The 'fail fast' mantra is often meant to encourage experimentation and reduce sunk-cost bias. It works when you're testing small ideas inside a validated framework. But when applied to entire product directions or business models, it can lead founders to give up before reaching real insight. Markets with longer sales cycles (e.g., B2B, healthcare, climate tech) don't reveal signals quickly. In these cases, what looks like 'failure' might just be a slow ramp. Giving up too soon wastes the chance to learn something valuable. For example, Airbnb struggled for years with low traction, relying on unconventional growth hacks and continuous iteration. If they had 'failed fast,' they would have quit before figuring out how to build trust at scale in a peer-to-peer model. The better version of this saying is to iterate fast. Fail cheap. But don't exit too early if you're solving a real problem. 3. 'Hire 10x Engineers' Can Stall Your Hiring Entirely The idea that a small number of great engineers are more productive than large teams of average ones is grounded in truth. But turning that into a strict hiring philosophy where only mythical '10x' candidates are acceptable can paralyze early hiring or foster toxic team dynamics. Also, the definition of a '10x' hire varies by context. In a startup, a strong generalist who can write decent code, talk to customers, and help set up infrastructure may be far more valuable than a brilliant algorithm designer who's uninterested in messy real-world problems. For example, Shopify's early team included product-minded engineers who focused on practical needs, not just technical brilliance. Their execution ability across a messy retail stack helped them scale fast. In startups, you should usually hire for adaptability, alignment, and ownership. A cohesive, good-enough team often beats a scattered group of 'stars.' 4. 'Your Startup Should Be A Rocket Ship' Can Lead To Premature Scaling Ambition is necessary. But the 'rocket ship' metaphor encourages companies to hire fast, raise more than needed, and go for growth before their systems, product, or customer base are ready. According to CB Insights, 70% of startups fail due to premature scaling - spending ahead of traction, growing headcount without process, or chasing growth metrics without retention. For example, Quibi, the short-form video platform, raised nearly $2 billion before launch, hired aggressively, and launched at massive scale. But with an unclear product-market fit, the platform collapsed within six months. 5. 'Talk To Users' Can Produce Misleading Signals Every founder hears 'talk to your users' or 'get out of the building' from the moment they enter the startup world. And it's good advice if you know what you're listening for. But talking to the wrong users, asking biased questions, or interpreting polite encouragement as validation can lead to false confidence. Early users often want to be helpful and will say what you want to hear, especially if they're friends or peers. Many failed social apps have strong feedback from early testers, but fail to scale because the use case isn't sticky or the need isn't strong enough to build a habit. You shouldn't just talk to users, but observe behavior, measure engagement, and look for patterns that confirm actual need.
TechCrunch
28 minutes ago
- TechCrunch
Google launches Doppl, a new app that lets you visualize how an outfit might look on you
Google is launching a new experimental app called Doppl that uses AI to visualize how different outfits might look on you, the company announced on Thursday. The app is available on iOS and Android in the U.S. Doppl is designed to let you virtually try on outfits on a digital version of yourself. The app works by first getting you to upload a full-body photo of yourself. From there, you can use photos or screenshots of different outfits to virtually try them on. These images could be a photo of an outfit you see at a thrift store or on a friend, or even a screenshot of an outfit you see while scrolling through social media. Once you select an outfit that you want to visualize, Doppl will create an image of a virtual version of yourself wearing the outfit. Doppl can also take these static images and convert them into AI-generated videos so you can get a better sense of how the outfit would look on you in real life. You can save your favorite looks and browse through all of your other virtual try-ons. Plus, you can also share your look with others. Image Credits:Google The tech giant says the new app builds on Google Shopping's recently launched capabilities that allow you to try on clothes virtually. By launching the functionality in a standalone app, Google is making the feature easier to access while allowing people to explore their style in a fun and interactive way, the company believes. It may also help Google collect more data on how apps like this could work, to aid its future efforts in the space. While Google has offered virtual try-on technology before, the earlier features focused on showing items on a diverse range of models' bodies. With Doppl, the company is letting you try clothes on an animated version of your own body. 'We hope Doppl helps you explore your style in new and exciting ways,' Google said in a blog post. 'As a Google Labs experiment, Doppl is in its early days and it might not always get things right. Fit, appearance and clothing details may not always be accurate.' Techcrunch event Save $200+ on your TechCrunch All Stage pass Build smarter. Scale faster. Connect deeper. Join visionaries from Precursor Ventures, NEA, Index Ventures, Underscore VC, and beyond for a day packed with strategies, workshops, and meaningful connections. Save $200+ on your TechCrunch All Stage pass Build smarter. Scale faster. Connect deeper. Join visionaries from Precursor Ventures, NEA, Index Ventures, Underscore VC, and beyond for a day packed with strategies, workshops, and meaningful connections. Boston, MA | REGISTER NOW Given that the Doppl is an experimental launch, it's unknown when or if Google plans to bring the app to additional regions.
Android Authority
28 minutes ago
- Android Authority
Google's new and improved Ask Photos is back on track — and rolling out now!
Google I/O TL;DR Ask Photos is a Google Photos feature designed to answer questions about the contents of your photo library. While announced at Google I/O 2024, access over the past year has been extremely limited. Earlier this month we heard the rollout was paused while Google worked to address a few issues, but today it's back on track. Over the past couple years, AI has been changing the way we do lots of things, and while there is absolutely no shortage of controversial uses of the tech, there are also some places where employing it just makes all the sense in the world. Back at I/O 2024, Google announced one of those really clever-sounding use cases, with Ask Photos for Google Photos promising to make your gigantic archive of pics actually useful by letting you search through it using natural language. After a very bumpy year, it finally sounds like Ask Photos is now ready to live up to its potential. Following that introduction, we caught an early look at Ask Photos last summer, but even by the time fall rolled around access was still extremely limited, with Google only testing Ask Photos with a small group of users. By 2025, it was starting to become clear that something was wrong, and earlier this month, Google Photos Product Manager Jamie Aspinall confirmed that the rollout of Ask Photos has been paused. Aspinall was surprisingly candid about the struggles the service faced, admitting failures 'in terms of latency, quality and ux.' While that was unfortunate to hear, he also teased that a much-improved iteration of Ask Photos was just around the corner. While Aspinall predicted we'd see Ask Photos make its return in about two weeks, it's been just over three now, and today Google confirms that Ask Photos is back on track. As of now, access is expanding beyond those early testers, and Google says that 'more' eligible Photos users in the US should be getting access. That suggest to us that Google's still being a little cautious with deployment, but hopefully everyone who's interested gets a chance to try it soon. To be eligible to use Ask Photos you need to be an English-speaking user in the US who has Face Groups toggled on. We also hear about some new functionality, and it sounds like Google cooked this up specifically to address some of the speed concerns we'd seen expressed about Ask Photos earlier. The app's basically combining classic search with this Gemini-powered upgrade to initially surface some broadly relevant results, while Gemini continues to crunch your query in the background. That means you get to see at least some pics straight away, rather than feeling like you're waiting forever on Ask Photos. Once you're able to give it a try for yourself, let us know in the comments if this was worth waiting for. Got a tip? Talk to us! Email our staff at Email our staff at news@ . You can stay anonymous or get credit for the info, it's your choice.
