Artificial intelligence may once have been dismissed as something out of a sci-fi novel, but the latest data from TourConnect AI shows it’s now very much at the booking desk.
The Sydney-based firm, which has been tinkering with travel-focused AI since before it was fashionable, has unveiled fresh statistics from its ItineraryAssist Copilot. The news is enough to make even the most sceptical Destination Management Company (DMC) manager raise an eyebrow.
In just 12 months, the system’s first-attempt accuracy in selecting the right hotel, transfer or service has leapt from a respectable 77.8% to a rather impressive 92.4%. That means more often than not, what the AI suggests goes straight into the client’s final itinerary, and no red pen is required.
Real Trips, Real Numbers
This isn’t marketing puffery. The figures stem from a rigorous analysis of 103,000 service line items across more than 7,000 genuine FIT itinerary and quote requests between July 2024 and July 2025.
In plain English, actual DMC teams fed in the requests using the Copilot in real-world conditions and the AI largely got it right.
Michael Herrmann, CEO and Founder of TourConnect AI, quickly acknowledges that this wasn’t simply the handiwork of clever coding. “Our job is to help DMCs build accurate itineraries, and we do it by paying close attention to detail, reliability, and our customers,” Herrmann noted.
That “customers” part is key. Much of the AI’s accuracy boost has come from direct feedback loops with DMC operations staff who aren’t shy about pointing out what works, and what doesn’t, when a machine is meddling with their day-to-day quoting.
What It Means for the Industry
DMCs have been racing against the clock for years, trying to respond to client requests before rivals swoop in. With the Copilot proving its worth, many DMCs report that quote turnaround times have shortened considerably.
Even better, no DMC that has onboarded the Copilot has since abandoned it. That suggests that confidence is running high and perhaps relief, too, given these teams’ endless administrative grind.
It also means AI is no longer a futuristic concept or optional extra. If a narrowly defined Copilot can perform at this level in 2025, the implications for the broader B2B travel ecosystem are profound.
Aiming Higher Still
TourConnect AI isn’t planning to rest on its laurels. The company’s next milestone is achieving a 95% acceptance rate for first-draft itineraries. And one newly released feature, Automated Product Rules (launched in August 2025), could be the ace up its sleeve.
This feature is a lifesaver for mid-sized DMCs that don’t have the resources to fuss with complex configuration. It allows the AI to learn from every human tweak, such as choosing the preferred transfer company, adjusting room categories, or selecting the right rental car class, and automatically builds rules tailored to that DMC.
The upshot is fewer headaches, less setup, and an AI that feels more like a seasoned colleague than a green intern.
Looking Ahead
Herrmann and his team remain focused on a simple mission: enabling DMCs to customise at scale. More features are already in the pipeline, and DMCs are intended to be kept firmly in the driver’s seat while the Copilot handles the heavy lifting.
For an industry that thrives on speed, precision, and personalisation, these latest figures show AI isn’t just dabbling in travel. It’s checking in, unpacking its bags, and very possibly extending its stay.
By Yves Thomas
