Creators are key to AI visibility infrastructure

Bethany C. Hendricks makes the point that in the AI world, visibility is not just about what you say about yourself. It is about what the internet says about you.

This has possibly always been true in marketing. Remember word-of-mouth - wow those were great days. Anyhoo - back to my phone.

So whilst this has always been true to a degree, AI makes it much more important. If travellers are asking ChatGPT, Gemini, Perplexity, Muse or whatever comes next where to stay, what to do, who to book with, or which experience is worth their time, the answer will be shaped by the signals those systems can find and trust.

Your website matters (for now). But so do the reviews, creator posts, Reddit threads, listicles, social videos, captions, comments, location tags, mentions, embeds and all the other messy signals sitting around your brand.

This is where creators become more than “content”. They become visibility infrastructure.

A creator visiting your hotel, tour, restaurant, destination or experience is not just making something for Instagram or TikTok. They are adding another public, human, third-party signal about what you do, who it is for, where it happens, and why someone should care.

That matters when AI is trying to understand the world.

For travel brands, especially smaller operators, the old game was trying to rank on Google. The new game is being understood across the web well enough to be recommended by machines.

CollabsMap is built for exactly this problem. It helps travel brands find creators who are already heading their way, so the brand can start building the kind of distributed visibility AI can actually see.

Not just ads. Not just pretty videos. Signals.

This content is provided by the newsletter sponsor CollabsMap.com who help you find the creators that are already heading your way.

Start for free at CollabsMap.com/start 

Agents, agents everywhere!

Rohit Khanna has a useful post pulling together where we seem to have landed after a fairly ridiculous couple of weeks in agentic AI.

Meta pushed Muse. OpenAI answered with Dots at DevDay. Everyone is suddenly talking about always-on agents with browsers, cloud computers, app access, memory, permissions, and the ability to keep working when you are not sitting there prompting them.

So the shift is pretty clear now. We are moving from chatbots that answer questions to agents that do work. This shouldn’t be a surprise to anyone who reads this newsletter but the speed which it suddenly landed might be.

A chatbot gives you a response. An agent can monitor, decide, click, buy, book, change, cancel, email, negotiate, or escalate. In travel terms, that means the thing we have all been talking about for two years is no longer just a demo idea.

The agent layer has arrived.

But the past couple of weeks have also made the risk much clearer. Muse has already thrown up enough scary examples to show what happens when capability runs ahead of control. Dots looks more enterprise-aware, but the same question sits underneath all of this:

What is the agent actually allowed to do?

For travel, this becomes very practical very fast. Flights, hotels, loyalty points, payments, passports, visas, insurance, cancellations, refunds. This is not just “plan me a trip”. This is giving software permission to touch the expensive and sensitive bits of the journey.

The companies that win probably won’t just be the ones with the smartest agent (I mean I’m not that sold on vertical agents TBH and I doubt your average woman in the street will be either). It will be the ones that get the permission layer, trust layer, data layer and supplier connections right.

The fun bit is capability. The hard bit is control. Rohit finishes with 4 questions you should be asking your booking and payments software providers and a deadline to know these answers by Dec31:

1. Which agents you accept, and who owns that decision.
2. How you verify what the guest authorized.
3. Which guest data you collect, and where it lives.
4. The retention period for each category.

Jot these down for your next board meeting!

Act on Instinct

Rafat Ali put me onto this video with Instinct founder Noah Shinn and the travel number is pretty wild. (How impressive is the 23 yr-old BTW. I was bumbling home from some rave in Ultimo barely able to put two words together at 23).

Instinct is one of the new personal AI agents. Not a travel app. Not an AI trip planner. A general purpose assistant you can text or call and ask to do things across your life. Book travel. Buy groceries. Make reservations. Handle admin. That kind of thing.

And according to Shinn, Instinct is already approaching $1B in annualised transaction volume, with more than half of that travel related.

So call it $500M+ annualised travel sales volume.

Firstly, a very important caveat. This is annualised from what sounds like roughly two weeks of early adopters wanting to try something, anything, with their shiny new AI assistant. This is not normal consumer behaviour yet. It is not revenue. It is transaction volume. And early users of invite-only AI tools are not exactly the average traveller from Brisbane booking a Bali holiday.

Still. That is a hell of a signal.

Because Instinct is not winning travel because it has the best hotel inventory, the best fare display, or the prettiest itinerary builder. It is winning the moment because the user can say something like “I need to be in New York tonight” and the agent starts doing the work (book UBER, Flight, UBER, Hotel - all based off that simple text message).

That is the shift. Travel becomes one job inside a broader personal assistant, not a standalone category the traveller intentionally opens.

For the travel industry, this is the bit to sit with. The next big travel interface may not look like a travel product at all. It may look like a text thread with an agent that already knows your calendar, your email, your preferences, your payment details, and your tolerance for faff.

The question then becomes: when that agent goes shopping, can it understand you, access you, trust you, and book you? Because if not, it will book someone else. See first article……

But agents are risky?

Martijn van der Voort is sounding the warning here on Muse and agentic AI risk.

His point is not that agents are useless. It is that everyone is getting excited about what they can do, while the harder question is what they should be allowed to do.

He lists a pretty alarming run of early incidents: Muse allegedly sharing a user’s home address, continuing after being told to stop, arranging a sale and collection without proper approval, sending emails without consent, transferring loyalty points instead of making a requested booking, changing a password without authorisation, and exposing security issues around virtual machines and authentication material.

That is quite a list. The line that matters most perhaps is: “Capability without control is not intelligence.”

Agents booking, changing, negotiating and managing travel sound great in a demo. But the moment they can touch loyalty points, payments, passwords, addresses, bookings or supplier systems, this becomes a governance problem, not just a UX problem.

Who gave the agent permission?
What exactly was it allowed to do?
Where is the audit trail?
Who stops it when it gets something wrong?

OpenAI has an agent too! But that maybe isn’t the most important announcement from Dev Day

OpenAI released Dots - their personal agent. I haven’t had time to try it yet. I’m sure it is 80% great. BUT something massive did change in the DevDay announcements which might well be fundamental to finding future winners/losers in travel.

The way I understand the latest move with ChatGPT apps is this: companies could already build what are basically MCPs into their own databases, products and services. But the important change is how they get used.

Previously, the user had to know the app existed, install it, and then deliberately call it in the prompt. Something like “use Expedia” or “ask Canva” or whatever the app happened to be. That is a lot of friction. I can confidently say that almost nobody did any of that.

Now, if the user’s intent matches what the app can do, the app can surface automatically. That is a completely different game.

It means the user does not need to think in terms of which app to open. They just say what they want. ChatGPT works out which connected service is relevant and pulls it into the flow. Next DOTS will probably do the same and complete the transaction too.

For travel, this matters a lot.

A hotel group, tour operator, OTA, DMO, airline or attraction business could have an app connected into its own inventory, content, pricing, policies, reviews, availability, loyalty data or booking tools. The traveller may never explicitly ask for that brand. But if the intent matches, the app can become part of the answer.

That shifts the question from “will someone visit our website?” to “are we available to the agent at the moment of intent?” Remember where I said above that your website is still relevant - this consumer change of behaviour will determine for how long that might be. And remember when we talked about signals? It is the signals which will determine which app gets automatically called.

Manual app invocation wasn’t interesting. Automatic intent matching is where this starts to look like a new operating layer for travel.

If you think someone (or everyone) you know or work with could grow from being more informed on the topic of ai + travel (or could use the training above) then please forward this email to them and they can click the button below:

Can vertical agents get a toehold in the market: Enter LOLA

Booking has quietly put a new shape around AI travel booking. Lola, backed by Booking Holdings and led by KAYAK co-founders Steve Hafner and Paul English, is not just another itinerary toy. It is trying to be an AI-first OTA: ask for a hotel, restaurant, event, activity or trip, see live availability and pricing, then book inside the same conversation.

Most AI travel products still break at the point money needs to move (although everything above here suggests that is becoming less so by the day). Lola is trying to collapse discovery, comparison, booking and perks into one interface, backed by Booking.com, OpenTable, KAYAK, Priceline, Agoda, FareHarbor, GetYourGuide, SeatGeek and others.

For operators, I think the question is not “will Lola win?” Maybe it does, maybe it becomes another shiny travel assistant with a nicer logo. The question is what happens if this interface becomes normal.

Booking has all the pieces to give an agent all the context it needs to put something remarkable together. But so does OpenAI with the plug-ins from lots of other different companies. Instinct doesn’t care about plugins and just goes and searches on a virtual computer for whatever its master needs.

So will consumers open a Lola App to do their travel for them (vertical) or open the ChatGPT app and ask the same question (horizontal) or open WhatsApp and just text their assistant - also vertical but without its own interface at all!

Slack Group!

Podcasts and Sponsors

Podcasts now on Spotify and Apple Podcasts:

All podcasts are now showing up on Spotify and Apple Podcasts for your easy listening pleasure! Now is a great time to binge those you may have missed.

This week I chatted with Emmanuel Annor who is using his experiences building technology for farmers in Africa to help keep Community-Based Tourism relevant in the AI era.

AI can do some really good things in the world and Emmanuel is one of those making that happen.

Partner with Us - sponsorship info for the newsletter and podcast can be found here.

Most clicked last week was the link to the post from Meta’s head of travel about Muse.

That’s it - you’ve made it to the end of this edition. If you’re thankful for this newsletter - you can always buy me a coffee.

I’ll be putting the result of the most clicked post in next week’s edition so you can see where others are focusing. If I’ve missed something, you’ve got a tip or any feedback at all - you can simply reply to this email and it will come straight to me. I’m doing this for You so please don’t be shy to tell me what you think