The majority of questions guests ask robots are way different to the ones they ask humans.
Quote from Akemi Tsunagawa, founder and CEO of Bespoke, in an article of PhocusWire this week.
Bots, bottled water and search - a varied snapshot of industry issues in 2018
"Test and learn" is a common phrase these days, littered throughout speeches and interviews with executives in the industry who want to portray how seriously they take their audiences and the progressive development of web services.
Some of it is based on assumptions - but most is data-led, in an attempt to figure out what works and what doesn't when deployed in front of consumers.
The results can often seem small to outsiders, yet a tweak to the size of a button, what tone within the brand colors are displayed in certain parts of a website or the wording used in so-called urgency marketing can increase conversions by a fraction of a percentage.
This, will all know, can mean an uptick - for large players - of millions of dollars over a quarter or 12-month period.
These tried and tested methods with web services do not necessarily apply to brand-new concepts, such as interactive tools in the guise of artificial intelligence-led interfaces.
"Test and learn," in this environment, is still fairly new as a concept, and assumptions are that data is required to understand what resonates and has an impact on travelers.
Tsunagawa's example of the initial failings of its bot is a prime example.
Consumers like shiny new toys, yet the interactions they have with them are based on the relative quirkiness of the device or service - rather than perhaps what the product is aimed at.
As Tsunagawa went on to say: "The majority of questions guests ask robots are way different to the ones they ask humans."
This is where new tech is right now - balancing a fine line between capturing the imagination of the traveler and being a genuinely useful tool to help improve or assist in the experience of a trip, or the customer's booking path.
There will plenty more mishaps and wrong assumptions along the way, this much is clear to anyone who has worked with integrating fairly outer-worldly ideas to an unsuspecting customer base.
The key is to not give up.
Commerce, human behavior and, lest we forget, travel itself are evolving faster than at any time in the history of the industry.
To dismiss the toys as just fads, when initial results are not as expected, is to ignore the inevitable change that will come before you.