Recently, there have been many news stories about the impact artificial intelligence has on search and its potential impact on the future of Google. The rise of generative AI in the last two years has allowed many searches to be conducted directly within AI models. Perplexity’s hybrid model has been available for some time now, and ChatGPT Search released last week could pose a significant threat against Google.
People value trustworthy sources. Brands and influencers give us more credibility. Results powered by the best source of knowledge in history — but backed up by trusted sources that we are already comfortable with — has the potential to shake things up further.
What are the next steps for travel companies? You should be aware of the changing dynamics of search and how to maintain online traffic from new sources.
Large language models [LLMs] They excel at providing general business information, but struggle with specific and current business information. These models have been pre-trained with vast datasets like the internet but do not continually learn or update facts. Unlike LLMs, tools like ChatGPT Search, Perplexity and Google’s Generative Search Results fetch live website data and summarize it in real-time, linking back to the sources — something LLMs cannot do directly.
If you assume that this new stage of generative AI won’t slow down, then you have two new objectives:
- LLM pre-training: to be discovered by the LLMs themselves, so that they can train on your content and serve up in response to prompts. This is a mystery. The most popular LLMs are trained on 12+ trillions of words. Even if you write 200 blog posts with 1,000 words each, it’s not enough to affect an LLM’s training.
- Generative AI search: to be discoverable by the LLM searching for data to enhance results. This is a place where you can make a significant impact. It’s good to know that if your pre-training is done correctly, you can also have a positive impact. It’s largely the same as standard search engine optimization.
Google Travel
Google Travel is a great product for the age of Ggenerative AI. It’s more about surfacing actual products (flights or hotels) than the company offering them. More and more products are appearing in standard search results. They may appear directly in a Google Travel search result or on a standard search results page. They also appear for a growing number of search terms.
LLMs struggle with live, structured product data. This is what generative AI searches rely on. Google Travel results have already been used to generate responses in Gemini, Google’s LLM. This includes hotels and flights. Gemini can, for instance, offer suggestions on hotels based directly on the Google Travel hotel feed if a searcher is looking for information regarding a trip to Paris. These are displayed with links to either the hotel’s website or an OTA. Google never conducts transactions.
Single source of Truth
There is currently no “go to” source for any of the other LLMs. This includes ChatGPT. Some companies will work with Google to provide this content. Most won’t — due to costs, conflicts, monopoly regulations, etc.
It’s difficult for us to pinpoint the exact source of all this traffic. ChatGPT is one of the most obvious sources. But Generative AI firms in the travel sector are also searching for this information. OTAs will happily share this information with you, as they already have it in their database. You need to be proactive if you hope to gain direct business.
You can assume that the information Google Travel and the major OTAs require for their listings are the same information you should prioritize on your website. These parameters and filters are there for a reason — because customers are looking for them.
So if you’re a hotel, having fluff like “Uncover the allure of [city] Hotel [name] as your sanctuary. Our boutique hotel seamlessly blends style and convenience to offer contemporary comfort and personalized services in the heart of the city. Immerse yourself whether you are traveling for business or pleasure. [city] vibrant culture from our urban oasis” is nice, but if you don’t include more on product details, the AI is not going to find you.
The long tail opportunity
This is similar to optimizing for long-tail keywords. Before generative AI, you might have been targeting very short keywords such as “San Diego Hotels.”
You might now appear in a generative search result when someone is having a discussion and asks, “Place to stay San Diego- walking distance from beach, pool suitable to kids ages 4 to 6 and open at night.” Balcony with ocean view. Restaurants within walking distance or on site. Rooms with desks suitable for working calls. Great Wi-Fi.”
The AI will understand the context. If you create product descriptions and website information, they will be well-matched for this search. They can be summarized and linked. Many of these variables can’t be sorted by filters. Many of them are probably not things you have ever considered optimizing for.
This is a great idea to consider. You can also find it in your Frequently Asked Questions. Previously, it was only meant for human consumption. You can add a great deal more information to this section in the future. This is more for the AI rather than humans.
Conclusion
OpenAI’s ChatGPT Search might not take a huge chunk out of Google, but when they release a product, it makes noise. It usually works.
All of these strategies involve some guesswork about what the future will bring. The good thing is that very few recommendations would not also be good if there were no changes. Future best practices are based on fundamental principles. These search fundamentals do not change. It’s all about doing the right thing, always keeping the end user in view.
Another thing is coming: the future user will not always be a human. It will be a robot that acts as a human. This is similar to an AI search engine that scans your content. The future agent comes pre-briefed about the goals and preferences its human. It will be expecting answers. If you do not provide the answers, after 0.001 seconds it will move onto your competitor.
