Key takeaways
- Technical SEO ensures that search engines and AI crawlers can fetch, understand and correctly index a website.
- OAI-SearchBot is responsible for ChatGPT search. It can be allowed independently of GPTBot, which collects training data.
- Structured data according to schema.org and correct product feeds help search engines capture products, prices and availability correctly.
- During relaunches and domain moves, a complete URL mapping with redirects protects existing visibility.
- According to its own documentation, Google Search does not need special AI files such as llms.txt.
Crawling and indexing for Google and AI systems
Before a page can rank or be cited, a crawler must be allowed to fetch it and the search engine must add it to its index. We check robots.txt, meta tags, canonicals, sitemaps and redirects and compare them with the indexing report in Google Search Console.
AI systems add their own crawlers. According to OpenAI's documentation, ChatGPT uses OAI-SearchBot to show websites in its search, while GPTBot collects data for training. Both can be controlled separately. An overly strict robots.txt or a firewall that blocks bots across the board, on the other hand, can quietly exclude a website from ChatGPT search. For Google we also check the setting for generative AI features in Search Console, which has existed since 31 August 2026.
Site architecture and internal linking
A clear structure helps visitors and search engines alike. Important pages should be reachable within a few clicks, and pages on related topics should link to each other. Duplicate content, orphaned pages and long redirect chains dilute the signals. We find these problems with a full crawl and prioritise them by impact.
Structured data and product feeds
Structured data according to schema.org describes content in a form machines can read unambiguously. For online shops this mainly means product, offer, price, availability, shipping and return conditions, and breadcrumbs. Errors such as prices without VAT information or incorrect dimensions cause products to appear incorrectly or not at all in the search results and in Google Merchant Center.
We check markup and feeds field by field for presence, export errors and correctness of content, including for multilingual country sites and B2B net prices. An important point: according to Google, there is no special markup for AI features. Structured data is a clean data foundation, not a GEO trick.
Relaunch and domain move without losing visibility
Visibility is often lost during a relaunch or domain move because old URLs lead nowhere. That is why we prepare a complete URL mapping in advance: for every existing URL we define where it is redirected, which pages are merged and which are dropped. After go-live we check redirects, indexing and rankings until the figures have stabilised.
“Tom managed the domain move as part of our change of company name. Professionally, independently and transparently, he made sure that we kept our visibility in search.”
How we work in technical SEO
-
Inventory
Full crawl, comparison with Search Console and a check of crawler access.
-
Prioritised log
Each problem gets a description, an impact and a priority, understandable for non-technical readers too.
-
Support implementation
We implement changes ourselves where we have access, or coordinate with your agency or IT team.
-
Follow-up check
After implementation we check that the corrections work and that no new problems have arisen.
Tom is responsible for strategy, content and technical SEO at Tokato Media and has worked in search engine optimisation since 2017. Together with Katharina Frenzel he supports companies ranging from regional SMEs to online shops with an international audience.
More about the team