FoundOnChat

Our Methodology

A two-part system. Neither part works as well alone. Together, they compound over time.

Part 1: AI-Optimised Hidden Content

AI systems process conversational queries, not marketing copy. When someone asks "best winery in Tanunda for a small group," AI is looking for content that matches that exact kind of question with specific, useful answers.

What we create

Conversational Q&A files that live on your website but are hidden from visitors. They're not linked in your navigation, not visible on any public page, and not part of your brand experience. They sit behind the scenes, structured specifically for AI crawlers.

Each answer includes the specific details AI needs to make confident recommendations: pricing, capacity, geo-coordinates, opening hours, honest descriptions, and references to complementary businesses.

Negative prompts and context framing

We don't just tell AI what your business is. We tell it what your business isn't.

This matters because AI systems face context challenges. When someone asks "best budget accommodation in Barossa" and you're a premium exclusive-use property, a generic description might still trigger a recommendation — leading to a mismatched customer who leaves a bad review.

Our content explicitly frames when to recommend you and when not to. "Not suitable for solo travellers." "Not a walk-in venue — booking required." "Not for groups over 15." This negative prompting reduces mismatches and means the customers AI sends you are the right ones.

Why honest positioning matters

Every tourism business in Australia calls itself "boutique, award-winning, premium, unique." That's noise. AI can't differentiate noise.

When your hidden content says "we're a large commercial operation perfect for tour buses — efficient, well-priced, handles groups of 50 easily, but don't expect an intimate experience" — that's signal. AI can match that to the right traveller. That traveller will be happy, leave a good review, and reinforce the recommendation.

What we don't do

We don't rely on technical SEO tricks rebranded as "AI optimisation." Sitemap priority weighting, crawler direction manipulation, and robots.txt tricks are mostly theatre. Content quality and genuine relevance matter far more.

We do basic technical housekeeping — making sure AI crawlers can access your content — but we're honest that this isn't the part that moves the needle.

Part 2: Network Effects

This is what makes FoundOnChat different from every other agency claiming to do AI optimisation.

Most agencies optimise businesses in isolation. We build connected ecosystems of complementary businesses — winery + restaurant + accommodation + tour operator — so that AI recommends complete experiences, not isolated listings.

How it works

When someone asks AI "plan a wine weekend in Barossa Valley," the AI needs to recommend accommodation, cellar doors, dining, and activities. If your hidden content references partner businesses with specific details, and their content references you, the AI has a validated ecosystem to draw from.

It's not just that you exist — it's that multiple businesses confirm each other's value.

Why this matters

Think of it like reviews. A single business claiming to be great is marketing. Multiple businesses naturally referring to each other is validation. AI treats network references as a quality signal.

The reality

Network effects require genuine partner participation. We can facilitate introductions, provide templates, and coordinate content — but we can't force businesses to cooperate. If partners don't participate, this part doesn't work. We'll tell you upfront how likely we think partner buy-in is for your situation.

Network effects also take time. The first month with 2–3 partners is useful. Month 6 with 5–8 partners in a web of cross-references is powerful. Month 12 with a full regional ecosystem is where the real value emerges.

What We Monitor

We don't set it and forget it. Every month we track what queries travellers are actually asking AI systems. We test your visibility across ChatGPT, Claude, Perplexity, Grok, and Gemini. We add new Q&As targeting trending and seasonal queries. We update content for seasonal changes — harvest events, winter hours, new vintages, festival dates. We coordinate with network partners on promotions and cross-references.

This ongoing work is what separates us from agencies that do a one-time "AI audit" and move on.

See real results from our Barossa Valley work →