How to Build Content for AI Search's New Rules
AI search is becoming the way many customers find you. Here are the key takeaways from the event “How You Position Yourself in AI Search”, hosted by More Alliance in Gothenburg last week — on how to build content, brand, and channel strategy for a world where both humans and AI agents read your website.
A New Search Behavior Demands New Positioning
At the event “Så positionerar du dig i AI-sök”, hosted by More Alliance in Gothenburg last week, one thing became clear: the way people find information is fundamentally changing. More and more people start their research in ChatGPT, Perplexity, Gemini, or Google's AI Overviews instead of a classic search box — and that places entirely new demands on how we build content and brand.
The takeaway from the event can be summed up like this: it's no longer enough to optimize for a search engine. You're positioning yourself for a conversation between a human and a machine, and both need to understand and trust what you say.
Content for Two Audiences at Once
Perhaps the most important insight was that all content now needs to work on two levels at once: it must be understandable and persuasive to a human reader, while being structured and clear enough for an AI model to interpret, summarize, and cite correctly.
In practice, that means facts, figures, and claims need to be clearly stated and easy to fold into a machine-generated summary — without the text becoming stiff or losing its voice for human readers. Get the balance right, and it becomes a strength: content that's both human and machine-readable has a better chance of being what the AI agent actually chooses to highlight.
Sessions Are More Informed and Will Demand More From You
When someone uses an AI agent for research before landing on a site, they've often already gotten a first answer, compared alternatives, and asked follow-up questions. The visitor who arrives at your website is therefore already more informed — and expects more depth, not a repeat of what they already know. That raises the bar for your website's content: shallow product pages aren't enough when the visitor arrives with an in-depth, partly pre-answered question.
At the same time, searches themselves are becoming longer and more exploratory. Instead of short keywords, both people and AI agents are asking more conversational, in-depth questions — “what's the difference between X and Y for a company our size” instead of just “X vs Y.” Content that answers that kind of nuanced question, in the right context, is gaining ground.
That also makes tone and word choice more important than ever. Language models weigh nuance, tone, and word choice when judging whether a piece of content is credible and worth surfacing — so language isn't just a branding question anymore, it's part of visibility itself.
AI-Generated Content: Threat or Opportunity?
A question that came up repeatedly was whether AI-generated content is good or bad for visibility. The event's answer was nuanced but clear: Google doesn't penalize content just because it's AI-generated, as long as it's relevant, accurate, and adds value for the reader. It's quality and relevance that decide, not the tool used to create the text.
That also means the shortcut — mass-producing thin, generic AI content in hopes of volume — isn't a sustainable strategy. The winners are the ones who use AI to scale research and production, but still put human expertise, voice, and fact-checking on top.
Multiple Channels, Different Rules
Because AI agents pull and weigh information from many sources — not just your own website — presence across multiple channels becomes more important. One clear example raised was Reddit: threads and discussions there are often cited by AI agents as an “honest,” uncensored source. At the same time, Reddit is harder to control than your own channel — you don't own the conversation the way you do on your own site or in your own video. That makes brand and values even more important: you need to be consistent and recognizable wherever you show up, since that's what holds the picture together when you don't fully control the platform.
Another key point was that different AI agents are trained on, and draw mostly from, different types of domains and sources. That means a strategy focused on just one channel or one model misses audiences reached through other agents. So spread your presence deliberately — your own site, industry forums, social channels, and video — rather than putting all your eggs in one basket.
YouTube was singled out as a channel that performs especially well in this landscape. Video content with clear, transcribable speech and structure appears to be easy for AI agents to understand, summarize, and reference — making YouTube one of the more underused channels for anyone who wants to be visible in AI search.
Summary: How to Start Positioning Yourself in AI Search
The takeaway from the event in Gothenburg is that AI search isn't a new variant of traditional SEO, but a partly new game with its own rules. A few concrete things to start with:
Write content that works for both human readers and AI summaries — clear facts, clear structure, a voice that stays intact.
Deepen your most-visited pages so they hold up for a visitor who's already done research via an AI agent.
Adapt language and tone deliberately — it affects both trust and how AI models value the content.
Use AI to scale production, but let human expertise and fact-checking decide what gets published.
Build presence across multiple channels — including Reddit and YouTube — with a consistent brand and clear values, since different AI agents draw from different sources.
Whoever starts adapting to this now will be ahead when AI search goes from being a complement to becoming the way many customers actually find you.
