Magic Pulpo Studio · 2026
Plenty of tools show where a brand is missing from AI answers. This one ships the fixes.
A studio venture: the implementation layer for AI search visibility. CrawlerPal audits how a brand shows up across ChatGPT, Claude, Perplexity, Gemini and Google AI Overviews, then ships the content and schema fixes, in every language the brand sells in. Built for the EU mid-market, live at crawlerpal.com in three languages.
The challenge
Mid-market brands in Europe are losing organic visibility to answers written by language models, and the GEO tooling that exists is measurement-heavy, US-centric and stops short of doing the work. A buyer asks ChatGPT and gets three competitors. A dashboard shows the gap. Nobody fixes it.
SEO specialists, content writers and a product manager came together to work out how brands get found now that people ask an AI instead of a search box. We researched it, tested it on our own work first, then ran it as a project for a client in Canada. It worked, so we made it a company.
The approach
CrawlerPal is a service first, deliberately. It audits each model (ChatGPT, Claude, Perplexity, Gemini and Google AI Overviews), traces the citation pathway from the buyer’s language to the sources the models cite, then ships the content, schema and crawler-access changes on top of whatever measurement tool the client already pays for. It becomes a product only once repeated delivery has shown which parts of the work repeat.
The studio built the whole surface: crawlerpal.com in three languages, the Instant Visibility Check (a five-question chat scored by Claude), the Snapshot audit workbench behind it, and the sprint playbooks. The full audit was free at first. It lost money on small deals, so the diagnosis is now the first phase of the paid Implementation Sprint and only the score stays free.
- 01
Scope the prompts
We map the buyers to about 20 prompts in their languages, and the client approves them before sampling starts.
- 02
Sample and diagnose
Each prompt runs three times on ChatGPT, Perplexity, Claude and Google AI Overviews, and the result is a ranked 90-day plan.
- 03
Ship the fixes
Content rewrites, citation outreach, schema and AI-crawler access, measured against the day-5 baseline.
The rule we learned
We first gave the full audit away free. It lost money on small deals, so the diagnosis is now the first phase of the paid Sprint, and only the score stays free.
The work
The results
The site and the visibility check are live and verified in production. The first ChatGPT baseline for an agency client sampled 28 prompts: the brand was mentioned in 19% of answers and cited in 15%. The next gate is the first paying client on the Implementation Sprint, from €6,500 per market. There is no revenue to report yet, so none is claimed.
What we delivered
- crawlerpal.com in English, French and Spanish
- The Instant Visibility Check: five questions, scored by Claude
- The Snapshot audit workbench: about 20 prompts across five engines, read by an analyst before the client sees it
- The Implementation Sprint, scoped per market
- Vertical playbooks for the first sectors
- A 30-second film
What it runs on
Claude
scores the Instant Visibility Check
Supabase
audits and leads
Five AI engines
ChatGPT, Perplexity, Claude and Google AI Overviews, sampled per prompt
Related projects
All work
We apply the same thinking to every engagement.
Book a 30-minute strategy call. We'll look at your challenge and say whether we're the team for it.