This week my team hopped on a call with one of the heaviest users of our agentic AI. We expected a certain kind of advisor but instead found ourselves on a Zoom call with a man in his 80s, beaming. He showed us how he was selling entire trips with a couple of sentences to the AI. He had all but stopped booking. His host agency assumed he had retired. He is now one of the top producers in his network.
That is what technology can do for the travel industry. It lets an 80 year old keep doing the work he loves. The same week, an agency owner told me they had freed up so much time on payouts that they could run an extra payroll cycle, putting money in their advisors' pockets sooner.
Hosts have always competed for advisors on training, education, community, commission splits, marketing, and culture. That competitive landscape barely moved for a decade. This week it became clear how fast it is moving now.
In the past week alone Jacques Ledbetter asked in Luxury Travel Advisor whether agencies should build their own technology or buy it. PhocusWire argued agentic AI is about to change how travel advisors work. And Fora raised 60 million dollars at a 1 billion dollar valuation, with the stated purpose of investing heavily in its AI tool.
Technology is the new battlefield, and the agencies that win the next decade will pair the best people with the best technology. I am calling this the tech enabled agency era. My goal with this post is to give agencies and hosts a mental model of how this new era will change the landscape.
This has happened before
Before we get into travel it’s worth noting that none of this is unique to travel. In 2011, Marc Andreessen wrote the seminal post that software was eating the world, and over the next decade it did. Everything from retail to music to finance was transformed. That transformation is coming to travel.
Travel advising held out longer than most because the work was so heavily based on unstructured data. The language (emails, PDFs, calls, etc) that advisors worked in daily was difficult for technology to work with. Large language models have changed that. For the first time, technology can take on the admin work and free up an advisor to do the uniquely human things that differentiates working with them from an online booking site.
When an industry hits that point, the winners almost always make the same move. No trading desk builds its own market data terminal. They rent Bloomberg and compete like hell on top of it, because no single company could build it and it would be crazy to try. Bloomberg was not just expensive to build. It required the best technologists in the world, and there are only so many of those to go around.
How to Build an enduring business in the Tech Enabled Agency Era
So here is the decision in front of every agency owner. Not build or buy. Build your own technology, or run your business on infrastructure built for the whole industry and owned by no one who competes with you.
Path one: build it yourself
There are two ways to build your own. Vibe code it, or hire a team to build it properly. Most agencies do not employ engineers, so for most owners the real question is the first one. Neither is a good option for a serious agency.
The vibe coding route
Credit where it is due. AI and a wave of app builders let people build software without knowing how to code, and that is real. I have a degree in Computer Engineering but have built tools for Tern I never could have written by hand using AI.
But AI is good at a specific kind of thing. The distinction is about stakes. If a tool could leak, break, or get wiped tomorrow and none of that would cripple you, vibe code it all day. Anything with sensitive client data or a system you run your business on is the opposite.
False Confidence Leads to Security Breaches
The risk is false confidence. AI will help a savvy engineer build something solid. It will just as happily help someone who knows the travel industry cold but has never thought about a database schema, a permission boundary, or a failure mode. That person gets a tool that feels like it works and is actually a house of cards.
You run one migration, answer one prompt wrong, and delete data you can never recover. You build something to store client credit card details and never think to set up a tokenization service to keep you PCI compliant, because that is not your world. Now you are holding raw card data you were never allowed to store. AI builds what you asked for, not what you forgot to ask for.
This is not hyperbole. A recent study by cybersecurity firm Redaccess analyzed roughly 5,000 apps built for corporate purposes. Of those, 40% exposed sensitive data including medical information, financial data, company secrets, etc. These leaks generally happen when environment variables aren’t set correctly, access controls are not properly thought about.
And even when it works, you are not done. Shipping is the cheap part. The expensive part is the years after, as everything around the tool changes. At Tern, we have over 4,000 feature requests in our feedback tool.
You ship something, move on, and then every request routes back to you, because you are the only one who understands how it was built. When technology is not your core business, that load only compounds.
The hire-a-team route
A great travel advisor does not sell twice what an average one sells. They sell orders of magnitudes more. Engineering has the same curve. The best builders are not a little better than the rest. They are a different category, and there are not many of them. You are competing with companies that are downsizing so they can pay these top engineers $1M a year.
At Tern, we’ve been in technology for over a decade. We’ve worked side by side with some of the best engineers in the world building multibillion dollar companies. Let me give you a sense of our hiring pipeline. For our AI & Data Engineering role we have gotten 3,450 applications since the job opened in late June. Of those, we interviewed 45 and expect to extend 1 offer.
When I was first getting into travel I was sharing learnings with a travel advisor who was building their own tech platform. They asked a question that is worth sitting with if you’re considering this route: “I’m not technical. How should I assess the technical capabilities of a potential technical co-founder”. The best answer I had was find an engineer you really respect and have them interview them.
If you are considering building a team, you need someone who knows how to sift through the 3,450 applications to hire the best. You can’t hire mediocre because then you’ll get mediocre technology and the only way to stay relevant in the tech enabled agency era is to compete with the gold standard.
The ROI of the best engineers is compounding in the AI era. Since January, AI has made our engineers about 300% more productive with no drop in quality. A world class engineer gets far more out of the same tool than an average one. The gap widens. Winning on technology takes the best talent empowered with AI, and that is expensive.
Large Hosts Agencies have tried this already.
Travel Planners International proved the point in public. They did not vibe code. They partnered with an external firm to build a real back office system called Suitcase. They ran it for 16 months across more than 6,000 advisors, and in June announced they were retiring it for Tern. Their CEO said it plainly. They built Suitcase because they believed in the vision, and "when it didn't fully deliver, we made the call to move on."
TPI is a great operator. They partnered with a third party that had a team of engineers. The tools that agencies need are enormously complex and not well understood by the outside world. Our PMs and engineers have spent hundreds of hours onsite learning the requirements, the edge cases, and the opportunities modern tools present. Tern can do this because we can amortize the cost across many agencies.
Path two: run on infrastructure built for everyone
Every summer my town hosts Art in the Park. It’s one of Michigan’s largest art fairs and my wife and I have made a tradition of walking through every part of the fair each summer.
Imagine talking to a local artist. They suggest that they want to sell their ceramics online so people who see them at art fairs can make a purchase when they get home. They say they’re thinking about hiring a few engineers and designers to build a custom website and payment platform. They expect it to cost just under $500K.
That would be insane. The economics would never work.
That’s why Shopify has become one of the biggest success stories in ecommerce. They’re now challenging Amazon in a way that was unthinkable not long ago. Shopify’s mission is simple: make commerce better for everyone. They empower any retailer to build a business online by offering everything a seller needs in one place. That artist can set up a shop in a few minutes and start selling their craft in a way that was never before possible.
At Tern, we are building Shopify for the travel industry. Thousands of agencies already run on Tern, and nearly $2B in travel flows through Tern in a year, double a year ago, across around 12,000 advisors.
Everything that an advisor needs to run their business lives in Tern whether they’re just starting out or they’re in their eighth decade. As an advisor, you maintain ownership and control of your business data so you can make the decisions that are best for you.
For a long time, making the leap from an advisor to an agency owner felt overwhelming. Lauren, from modern travel, spent years dreaming of starting her own agency but the back office logistics felt too overwhelming. Tern enabled Lauren to start Modern Travel, which is rapidly becoming one of Canada’s leading travel agencies.
That infrastructure scales. Large hosts like Travel Planners International and Uniglobe are among the first to take that further and offer every advisor and host agency owner an entire operating system with Tern.
One thing decides the rest. Speed.
The tools you need are not finished. They will keep changing as fast as AI changes. A partner that keeps pace is worth everything. One that stalls leaves you stranded on software you cannot fix yourself.
At Tern, speed is what we have spent a decade getting good at. At the start of this year we had almost no agentic AI. Today more than two-thirds of our advisors use it, built and shipped in a single quarter. Our 3 product squads ship to advisors every week, and every Thursday our head of product, Molly Johnson, posts a public log of what went out. Read it, then find anyone else in this space who will show you the same.
The platform also keeps growing under you. An advisor can already drop a cruise into Tern, ask for shore excursions matched to their clients, and book them without leaving the platform. Booking is expanding category by category, and the same tool you use to run your business will handle more of the trip every quarter. You are not buying what Tern is today. You are buying everything it becomes, without paying to build a line of it.
What this means for you
So if you own an agency, my honest read is this. You will compete on technology whether you want to or not. The only question is how. If you think you can out build a company whose whole reason for existing is travel technology, go for it, and know what it costs. If you would rather compete on what you are great at, we built Tern for you.
Host agencies should compete on what they have always been best at. Recruiting, training, teaching people to sell, managing risk, building community. Those are hard, human things, and no platform should take them from you. Let us carry the technology. You carry the relationships.
If you are watching all this and worried you cannot keep up, here is what I want you to know. You do not have to. That is our job. Tern has your back.