Services · Marketing
AI SEO
How to prepare a website so that ChatGPT, Gemini, Copilot, Perplexity and Claude discover it, read it, understand it, trust it and cite it. The goal is not just “to be on Google”, but to be the most useful source for a specific question.
What we take care of
- Site accessibility for AI crawlers
- Machine-readable data (schema, feed)
- Content that answers directly
- Trust signals and measurement
How AI gets to your site
A chain in which no link can be missing
If any link is missing, AI often cites someone else, even if your offer is better.
- Crawl
- Index
- Understanding
- Citation
- Click
- Conversion
Content
Keywords are no longer enough
More and more often users are not looking for “10 blue links”, they ask AI: “Which product is best for X?” AI cites the most useful answer, not the best store. It prefers a page that has:
- a clear answer in 1–2 sentences
- a comparison: criteria, who it is for, pros and cons
- frequently asked questions: how to choose, size, compatibility, delivery, warranty
- concrete data and specifications
That is why, alongside product pages, it pays to have category guides (e.g. “How to choose …”) that link internally to products.
Technical foundation
Six conditions for AI to cite you
Accessible to bots
AI crawlers can visit and read the site (crawl).
Sensibly indexed
Important pages are in the index, duplicate and unimportant ones are not.
Machine-readable
Structured data (schema) and a product feed.
Answer first
The content directly answers a specific question (answer-first).
Trust signals
Delivery, returns, contact details and verifiable information.
Measurement
You track whether AI sends traffic and conversions.
There is no single AI index
Each system reaches its sources differently
That is why you need a universal technical foundation and content that can be cited.
Google (Gemini, AI Overviews, AI Mode)
Largely based on Google’s search index and its guidelines for site owners.
ChatGPT (OpenAI)
OAI-SearchBot is key for appearing in ChatGPT search, so you must not block it in robots.txt.
Perplexity
It has its own crawler (PerplexityBot); changes in robots.txt are picked up with a delay of up to about 24 hours.
Claude (Anthropic)
It distinguishes between its search crawler (Claude-SearchBot) and fetching at a user’s request (Claude-User).
Example: WooCommerce
The 12 most common mistakes
- Filters create thousands of URLs and use up the crawl budget
- Wrong canonical, so categories and products compete with each other
- Product schema has no Offer data (price, stock)
- The feed has no GTIN or brand, or price and stock do not match
- Robots.txt accidentally blocks products or categories
- A CDN or firewall blocks bots (403/429)
- Content is too general, with no selection criteria or FAQ
- The site has no trust signals (delivery, returns, contact)
- Product descriptions are too short
- No internal links from guides to products
- Too many similar categories and tags (duplication)
- No routine for updating stock, prices and FAQ
The best place to start: pick 5–10 pages that already bring revenue and fix them in the order crawl → index → schema/feed → answer → trust → measurement.
AI Index Checklist – a 30-minute diagnostic
Download the free checklist and check in half an hour whether your site is ready for AI search engines. The checklist is in Slovenian.