Not long ago the first question a prospective guest asked was typed into Google. Now some of it starts in a chat window instead. Still a small share – single digits, on the best numbers available – but it quadrupled in a year, and it is the share that arrives already half-decided: which multi-day walk in New Zealand suits a moderately fit couple in March, with a small group and transfers included. One answer comes back. Three or four operators get named in it. Everyone else is invisible – not ranked low, absent.
That is a different competitive surface to search, and it does not respond to the same work. Ten blue links rewarded whoever earned the click. An answer rewards whoever the model can describe confidently and correctly. Confidence comes from consistency: the same facts about your operation, stated the same way, in enough places that the model treats them as settled.
If you want the mechanics rather than the story – how answer engines choose sources, and which of the advice going around is folklore – we set that out in what AEO is and how it differs from SEO. This piece is what it looked like on one account.
What actually changed.
Three things, and only the third is new. Search demand did not vanish – it split. High-intent commercial queries still convert on paid and organic. Research queries moved. And a handful of sources get named per topic rather than ten, so the gap between being in that handful and being outside it is much wider than the gap between position three and position eight ever was.
For an operator, the practical consequence is that your itinerary pages are no longer only sales pages. They are the reference an assistant reads when it decides whether you match the trip somebody just described.
“Ranking is a position. Being cited is a description. You can only fix the second by fixing what you publish about yourself.”
The four moves that mattered.
I ran this on New Zealand Trails over one quarter, alongside the paid and organic work rather than instead of it.
- Fixed the facts before the content.
One canonical set of numbers – departure dates, group size, distance per day, fitness rating, what is included – stated identically everywhere they appear, on site and off it.
- Rebuilt itinerary pages as references.
Day-by-day detail, terrain, conditions by month, who the trip suits and who it does not. Written to be quoted, not to be skimmed.
- Earned third-party corroboration.
Models trust agreement. We pursued the specific mentions and listings where the same facts are repeated by sources that are not us.
- Measured citation, not just traffic.
A monthly prompt set run across the main assistants, logging who is named and how accurately we are described. That log is the roadmap.
What it moved.
Assistant referrals are still a small share of sessions and I am not going to pretend otherwise. What I can say is that they arrive late in the decision, with the trip already half-specified. We did not have the volume to put a trustworthy conversion figure on that, so I am not going to invent one.
The same work also lifted conventional organic performance. That surprised the client and it should not have surprised us: clearer facts, better structure and real depth on each itinerary are what classical SEO has asked for all along. We did one piece of work and got paid for it twice.
Where to start this week.
The numbers at the foot of this page are the whole account over the year, not this quarter of work on its own. I am not claiming one caused the other.
Ask the three assistants your guests use to plan a trip like yours. Write down who gets named, and what each answer says about you – including the parts that are wrong. That list is your brief. Most operators find the first fix is not content at all: it is a fact about departures, group size or fitness level that three pages of their own site state three different ways.