Welcome to our May newsletter!
If you’re here, you already know search is changing.
Something is becoming clearer every month: AI visibility is not the same as brand visibility.
When we start conversations with new clients, we often find that they think their visibility is ‘fine’. They already appear when people search for their name, their category, or the service they offer. Maybe they don’t need to worry…
But the real opportunity (and the miss for so many) lies where people are still working out what problem they have, what their options are, and who they should trust to help them.
This is where AI search diverges from traditional search comes into its own.
This where many brands are still invisible, and where you can really stand out.
Below are three things we’re seeing right now, along with our take on what they mean in practice:
- AI search is exposing the middle-of-funnel gap
- AI visibility has an environmental cost
- Distribution matters more than ever
schedule
Read time:
short_text
Word count:
1: Being visible for your brand name is not the same as being useful in AI search
One pattern keeps coming up in our client audits.
A brand believes it has strong visibility because it appears when people search for the business directly, or when the query is already quite specific.
That might mean showing up for: “Is x a good holiday park?”, “Show me x product range”, “Where can I buy x service?”, or “Best x providers near me”
Yes, that visibility matters, but the customer already knows the brand, understands the category, or has decided what kind of solution they need. How hard can that be… and how many people are you missing who don’t know what they want yet?
The test is whether your brand appears when someone is still working out what they need.
- “How do I choose the right holiday home to buy for my family?”
- “How can I reduce digital transformation risk?”
- “What luxury handbag is as good quality as Hermes but within my price range?”
These decision-making questions are the heartland of AI search. They reflect people exploring, comparing, and scoping out their options.
OpenAI’s own research says around three-quarters of ChatGPT conversations focus on practical guidance, seeking information, and writing, with 49% of messages classified as ‘asking’, where people use ChatGPT as an advisor rather than simply as a task tool.
McKinsey also reports that AI-powered search is already being used across the consumer decision journey, with around 40–55% of consumers in major sectors using AI-based search to make purchasing decisions.
So what?
Don’t just ask whether AI can find you. Ask whether AI understands where your business fits when the customer is still shaping the question.
2: AI has real costs. Better content reduces wasted effort.
AI search brings huge opportunities, and it is easy to forget that those opportunities come with a very real environmental footprint.
The numbers around individual prompts can look surprisingly small. The typical Gemini text prompt consumes the same energy as watching about 9 seconds of TV or running your microwave for 1 second.
Sam Altman claims a single ChatGPT query uses ‘just’ 0.34 watt-hours of electricity and 0.000085 gallons of water, this is what a high-efficiency lightbulb uses in a couple of minutes, plus roughly one-fifteenth of a teaspoon of water.
So ok, maybe one prompt is not the issue. The issue is scale.
ChatGPT was reportedly processing more than 2.5 billion prompts a day in 2025. And that’s just one part of the picture. These prompt-level figures focus on the energy used to generate a response, not the full infrastructure around AI: model training, data centre cooling, hardware production, web crawling, storage, networking, and the constant expansion of the systems that keep these tools running.
It’s the full context that businesses need to understand.
AI is far from weightless. Every prompt, crawl, summary, repeated question, and low-quality answer sits inside a growing infrastructure of data centres, cooling systems, chips, networks, and energy supply.
Human costs are also very real. The fascinating MIT Media Lab’s ‘Your Brain on ChatGPT’ study raised concerns about ‘cognitive debt’ when people use AI writing tools heavily, while the wider debate around “dead internet theory” reflects growing unease about a web increasingly filled with bots, synthetic content, and low-trust signals.
What can we do?
We are not saying you should stop using AI, or pretend people won’t use it.
The answer is to make the web more useful.
- Clearer content means fewer repeated searches
- Better structured answers mean less wasted inference
- More authoritative information means fewer poor-quality summaries
- And stronger human-led content means the web remains somewhere worth learning from.
For us, this is where GEO and good content overlap.
The goal is not to create more stuff for AI to chew through. It is to create better content that helps people, and machines, get to the right answer faster.
3: Your website matters, but it is only one part of AI visibility
The web is dead? Not at all, your website is still an essential part of your toolkit.
But AI search is not only looking at your website.
AI search draws from a wide system of signals: YouTube, media coverage, third-party mentions, review platforms, forums, Wikipedia, structured HTML pages, and other sources all help it decide what is credible, useful, and worth citing.
OtterlyAI’s 2026 AI Citation Economy report analysed more than 1 million citations across ChatGPT, Perplexity, and Google AI Overviews, showing just how varied the citation landscape has become.
This is why distribution now matters so much.
The report makes it clear that AI visibility is not won on your website alone. Community platforms and Wikipedia dominate AI citations, and news and media represent 20–30% of citations depending on the platform.
Social and video platforms accounted for 5.54% of AI citations, with YouTube taking 31.8% of that social citation share, second only to Reddit.
Off-site authority is a core part of GEO. Brands need to show up strong beyond their own domain, through earned media, useful video, third-party mentions, community visibility, and credible reference-style content in the places AI systems already look for answers.
In a nutshell, your website is the foundation, but the whole system is much bigger.
To build AI visibility, brands need to think about where else their expertise appears.
That could mean:
- Publishing useful video content on YouTube
- Turning key expertise into press stories or digital PR
- Strengthening third-party mentions and citations
- Improving review and comparison visibility
- Building genuinely helpful website content in clean HTML
- Making sure owned, earned, and distributed content all tell the same story
In traditional search, your website was the destination. In AI search, your website is just one of the sources AI uses to decide if you are the answer.
Look out for the next edition, where we’ll continue to share where to focus if you want to stay visible to AI, while still delivering content that engages and delights the people you are trying to reach.
More soon.
Jo.
About the author
What we do
Frontier15 helps brands achieve visibility in the era of AI search, creating content strategies that make your expertise discoverable and trusted by both people and machines. Get in touch and ask us how we can map a GEO-informed content strategy that helps future-proof your brand.