Brand story meets machine readability

We work alongside The Storytellers Collective to combine two things that are usually done separately: a) refining how a brand communicates its story, and b) ensuring AI systems can find and utilise it. One without the other leaves a gap. A perfectly optimised site with nothing distinctive to say is discoverable but forgettable. A compelling brand narrative - locked behind a JavaScript framework - is invisible. 
However, get both of these right - and lean into the potentials of agentic AI agents ingest and utilise your brand channels - and you can start to shape how the AI powered future will discover and see your brand.

Case study: A well-presented brand, effectively invisible 

We recently completed an AI Compatibility Check-in for a London-based luxury brand: strong content, clear positioning, a solid Trustpilot profile, and recent editorial features in major industry outlets. By any conventional measure, a well-presented brand. 
Yet - when we examined what AI systems actually see when they visit the website, the answer was essentially nothing. The site is built on a modern headless framework that delivers content via JavaScript streaming. AI crawlers do not execute JavaScript. Every product description, brand story, FAQ, and pricing comparison was invisible. A carefully written website, effectively blank.

Three layers, not one 

The first is training influence: how a brand's public content shapes LLM perception over time. Consistent terminology, third-party mentions, and credible authority signals all matter here. Vague or unstructured content gets ignored; keyword stuffing actively reduces visibility. What works is specific, well-sourced material: statistics with attributions, expert quotations, and cited claims that give AI systems something concrete to extract and repeat.

The second is real-time retrieval. When a consumer asks Perplexity or ChatGPT to recommend the best cashmere jumper, the AI goes looking. Whether it finds your brand depends on whether your content is technically accessible, semantically structured, and useful to a system that cannot click, scroll, or execute JavaScript. Research from Carnegie Mellon has shown that even subtle changes in how consumers word a query can shift which brands get recommended by as much as 78%, which means your content needs to hold up across many variations of the same question, not just one.

The third is agent interactability: whether a shopping assistant or autonomous purchasing agent can navigate, evaluate, and transact with your products. This is further along the adoption curve, but whether Shopify's emerging protocol integrations can surface your products to a shopping agent next year depends on platform and data choices you make now.

Many brands, and evaluations, address only the first layer. The second is where most of the immediate value lies. The third is where early decisions will determine who gets selected and who gets skipped.

What we actually look at 

Our evaluation covers seven areas: 

1. Identity clarity: how well a brand communicates what it is and who it serves, in a form AI systems can actually extract.
2. Content structure: whether pages are organised for machine parsing, not just human reading.
3. Authority signals: the presence and accessibility of statistics, expert attribution, and social proof.
4. Technical infrastructure: schema markup, server-side rendering, and crawler configuration.
5. Narrative consistency: whether the brand tells the same story across every platform an AI might encounter.
6. Product information completeness: enough detail for an AI to construct a confident recommendation.
7. Agent readiness: whether product data is accessible to systems that need to act, not just read.

The pattern that recurs 

The raw content is often already there. What is usually missing is the structure and presentation that makes it work for machines: heading hierarchies that give AI systems clean chunks to extract, FAQ formats that map to how consumers actually query, product data complete enough for an agent to build a comparison. The technical barriers are frequently fixable. The content restructuring takes more thought, but it builds on what already exists rather than starting from scratch.

If you would like to understand where your brand stands, we are offering a limited number of AI Compatibility Check-ins to brands that want to get ahead of this. Get in touch.