Vidima AI

The problem it solves

The zero-click shift: "Ten blue links became one answer." Google gave your customers ten links; AI gives them one answer. If your business isn't in it, you don't exist for that buyer — and you'd never know. Hero line: "Your customers now ask AI. Are you the answer?"

Who it's for (ICP)

SMB owners and lean marketers without a dedicated marketing person, whose buyers now ask ChatGPT / Perplexity / Google AI Overviews before they buy. Sharp wedge: non-English / small-language markets, Bulgarian-first — businesses whose customers ask AI in their own language. (Beachhead customer: Sky-R, a B2B promo-products supplier in Bulgaria.)

What it does — the Scan → Fix → See loop

• Multi-engine scan: runs buyers' real purchase-intent prompts across 4 live AI engines — ChatGPT, Perplexity, Google AI Overview, Gemini — on a schedule, in the buyer's language, multi-sampled (because a single answer is a coin flip).
• Visibility score (0–100): how often and how favorably AI engines name your business, tracked week over week, with every answer kept as proof.
• Receipts: each score opens into the actual quoted AI answers — who was cited, and where you were missing.
• Share-of-voice face-off vs named competitors, overall and per engine, as a printable one-pager.
• Ranked, plain-language fix list ("work order"): on-site fixes you do + off-site fixes Vidima drafts (it finds the contact and writes the ask in the site's language). Built on five evidence-based levers: topical relevance, structured data, metadata & freshness, semantic HTML, third-party presence.
• Scheduled re-scans / proof loop: the trend shows what actually moved.
• Multilingual-first: language detection, local-first prompts, localized reports (French, German, Bulgarian, Hindi, Japanese, …) — not English-only.
• Plus: prompt/keyword discovery onboarding, and an AI-crawler access audit (robots.txt / Cloudflare check for the AI retrieval stack).