Welcome to our April newsletter
If you’re here, you already know search is changing.
AI search is moving from lists of links to direct answers. That shift changes how visibility works. It is no longer just about being present, but about being understood well enough to be selected.
The timeline is still playing out, but behaviour is already moving. The more useful question is whether your content is ready to meet it… without compromising the experience you create for your customers.
Below are three developments shaping that shift right now, along with our take on what they mean in practice:
- AI becomes a transaction layer
- Google shifts from ranking to interpretation
- The web adapts to AI — with mixed results
schedule
Read time:
short_text
Word count:
1: Shopify made AI discovery much more real for commerce brands
In January, Shopify announced its Universal Commerce Protocol, (UCP) developed with Google, enabling retailers to sell directly within AI-powered experiences like Google’s AI Mode and Gemini.
This takes AI from being a research tool to something much closer to a transaction layer.
With UCP, product content (such as descriptions, FAQs, structured data, etc) doesn’t just support customer decision making. It becomes instrumental in determining which products are discovered, compared and purchased within those AI-driven journeys.
Shopify is not alone here. commercetools is developing its own Agentic Commerce Protocol (ACP), enabling AI agents to discover and transact using structured product data. This marks another clear signal of the inevitable move towards AI-native commerce.
For retail brands, this is a material shift. As content becomes an integral cog in the buying process itself, clarity and structure matter far beyond marketing. Together they will help your content function almost like infrastructure.
2: Google is leaning harder into AI answers
Earlier this year, Google confirmed that Search is now powered more deeply by Gemini, expanding AI Overviews and introducing follow-up questions directly within results.
In its own guidance, Google is clear that inclusion in these AI experiences still depends on the fundamentals: content that is crawlable, indexable, and genuinely helpful to people.
What is changing is how that content is used. These systems are designed to interpret and synthesise information, not just rank it, which shifts visibility from a ranking exercise to something closer to an interpretation exercise.
If your content is vague or difficult to parse, it is less likely to be used. If it is clear, structured and specific, it stands a better chance of being surfaced when the question is asked.
3: The web is starting to adapt to AI consumption – but don’t go jumping without looking first
In February, Cloudflare introduced “Markdown for Agents”, a way of serving simpler, cleaner versions of web pages specifically for AI systems, making it easier for them to read and process content.
This means that, for example, a page with over 16,000 tokens in HTML drops to just over 3,000 in Markdown. This means a significant reduction in what AI needs to interpret.
Taken at face value it looks like this could be a simple move to gain visibility.
But recent testing from Frontier15 partners OtterlyAI revealed an interesting fact. In a controlled experiment, their HTML pages received 7.4% of AI bot visits over a two-week period, while equivalent Markdown versions received none. They weren’t cited by AI platforms at all.
For the details, checkout this post from Otterly founder Thomas Peham
Read Cloudflare’s announcement: https://blog.cloudflare.com/markdown-for-agents/
Our take on it is this: don’t rush at chasing new formats before knowing what actually gets used.
AI systems may evolve towards cleaner inputs, but right now, visibility is clearly still being driven by well-structured, high-quality HTML content.
Your priority shouldn’t be to rebuild everything in Markdown. Rather it’s to make sure what you already have is clear enough to be understood (by humans and machines), and strong enough to be referenced.
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.