For those outside of travel industry, GenAI+travel planning seems to be a no brainer? Just look at the number of GenAI travel planners that have popped up recently: (https://www.reddit.com/r/AI_travel_tips/comments/1edaay2/list_of_ai_travel_planners/). Similarly Perplexity, OpenAI and other GenAI companies keep diving into this area. However the situation is a lot more nuanced than it might look like.
Anyone planning to build an AI travel planner should keep an eye on the state of current Gen AI travel planners. Google seems to have given up on its plan to build Gen AI planner after 2024 keynote. Similarly other Gen AI travel planners are struggling to monetize. So is there are reason why a product so easy to build is so difficult to monetize?
In ecommerce we call this phenomena a 'high risk + high cost' situation. Users have a very high inertia when it comes to buying a high risk or a high cost product or service. This is why I would not buy a $1000 iphone from a random website despite large discount. This is why I would try coffee from a new coffee shop but not switch my barber (imagine walking around with bad hair cut for a month). Travel falls under both high risk and high cost scenario and as a result users are very reluctant to book hotels/tours unless the incentive and confidence signal are very strong.
The future of AI travel planners lies in a mix of human + GenAI based travel planning. At PhotoSpot, we are experimenting with the righ hand off point where GenAI hands over planning to a human expert. Most people think they can find 80-90% information on their own. Its the last 10% where they need confidence that their plan seems reasonable. That should not cost $300-400. With PhotoSpots, AI travel planner users can talk to a licensed tour guide for $30 and travel with confidence.
Top comments (0)