SEO and AI visibility. I fix, I do not promise.
I do SEO from the technical side: an audit of Core Web Vitals, crawling and indexing, schema.org structured data and preparing the site for AI answers - ChatGPT, Perplexity, Google AI Overviews. I do not promise rankings: I fix what measurably blocks visibility and show the effect in numbers. I work like an engineer - the diagnosis in writing, the implementation straight in the code. The quote is always concrete: amount, scope, date.
Technical SEO - how I work.
Technical audit - what measurably blocks visibility.
I start with a diagnosis in writing: what to fix, what to postpone, what to remove. Three layers, in this order:
Core Web Vitals and performance
LCP, INP and CLS are ranking factors - a slow site loses on mobile before anyone reaches the content. I diagnose with a network waterfall and a real measurement, then fix the bottlenecks: caching, images, theme and plugin bloat. The effect is measured before and after.
Crawling and indexing
Google will not index what it cannot reach. I check the sitemap, robots.txt, canonicals, duplicates, redirect chains and the errors in Search Console - and tidy the structure so the crawler spends its budget on the pages that sell.
Meta, headings and content architecture
Correct titles and descriptions, a logical heading hierarchy, internal linking and a sane URL structure. Boring? Maybe - but it is the foundation, and without it no content stands a chance.
Structured data and entity - so machines know who you are.
Search engines and AI models do not read a site the way a person does - they build a profile of the company out of it. That profile has to be complete and consistent:
Structured data with schema.org
Organization, Service, FAQPage, Review - the markup that tells Google and AI models who you are, what you do and for whom. Written into the code, not generated blindly by a plugin.
A consistent company entity
Company registration details, address and contact data identical in every place: on the site, in public registers and in external profiles. Search engines and AI models verify a company the way an office does - a mismatch in the data costs you trust.
Proof instead of declarations
Projects with numbers, reviews with a name under them, a clear answer to "who is behind this". That is the kind of content search engines and AI models quote - nobody quotes generalities.
How do you get visible in AI answers - ChatGPT, Perplexity, AI Overviews?
More and more customers ask AI before they open the classic Google results. If the models have nothing to quote about you, that traffic passes the company by without a trace in the statistics. Four things that genuinely move it:
Content a model can actually read
AI quotes pages that answer directly: the question in the heading, a concrete answer in the first paragraph, structured data underneath. I rebuild the key pages in that direction - without ruining them for people.
llms.txt and access for AI crawlers
An llms.txt manifest with an up-to-date summary of the offer and the content, correct headers for AI bots, content available without JavaScript wherever that is possible. Small details that decide whether ChatGPT and Perplexity have anything to quote at all.
Content parity - FAQ visible 1:1
What you declare in structured data has to exist on the page word for word. A gap between the schema and the content is the shortest road to losing trust and rich results.
Measurement instead of guesswork
I check what ChatGPT, Perplexity and Google AI Overviews actually say about the company and about buying-intent phrases - before the changes and after them. AI visibility is a process, but you know the starting point and the direction from the first report.
Migration without losing traffic - 301 maps before the switchover.
Changing the store engine, the domain or the URL structure is the moment companies most often lose their rankings. Not because they had to - because SEO was the clean-up after the migration instead of part of it:
A map of 301 redirects
Every old URL mapped to a new one, with no chains and no loops. This is the most common reason traffic disappears when a store moves or a structure changes - and the easiest one to avoid, as long as the map is written before the switchover, not after it.
Meta, canonicals and sitemap
I carry over the category structure, the titles and the descriptions, set the canonicals and generate a fresh sitemap - so that from day one Google knows what is what in the new structure.
Indexing under control
After the switchover I watch Search Console: index coverage, 404s, redirect statuses. Problems get caught in days, not once the traffic has already dropped.
Planning to move a whole store onto WooCommerce? I describe the full process - data migration and staging included - on the e-commerce stores page.
Would you rather read first?
Two pieces that show you what your site is actually standing on:
The same method, a measurable effect.
The public proof of that work on the technical foundation is Złota Faza: the performance audit pushed from 38 to 100, Speed Index from 4.5 s down to 0.3 s. That exact layer - performance, structure, data - is what also decides visibility in search and in AI.
"Krystian talked less and did more. He quoted it, gave a date, delivered. The site is up, Google sees it, customers come in." Michał Wacławik · Złota Faza · electrical installations
Before you ask.
I do not sell a monthly retainer "for SEO". Once you describe the site and the goal, you get a fixed bid in writing - the scope follows the problem: sometimes it is a one-off audit with the fixes implemented, sometimes a bigger job around a migration or AI visibility. You pay for concrete, measurable changes, not for reports.
No - and nobody honest can, because neither the contractor nor you controls the results. I guarantee something else: I will fix what measurably blocks visibility (performance, indexing, structured data, redirects), and you will see every change in numbers - in Search Console and in the before and after measurements.
I am a developer, so an audit ends with implementation - I put the recommendations into the code myself: from caching and images to schema.org and redirect maps. I do not buy links, I do not run link farms, and I say it straight when the problem is not technical.
More and more customers ask ChatGPT, Perplexity or Google AI Overviews "who do you recommend for X" before they ever open the classic results. If AI does not know you, or has nothing to quote, that traffic passes the company by without a trace in the statistics. I prepare the site so the models have something to quote: content that answers directly, structured data, llms.txt and a consistent company entity.
No, as long as SEO is part of the migration rather than the clean-up after it. I map every old URL to a new one, set the 301 redirects, carry over the meta and the category structure, generate a sitemap and watch indexing in Search Console after the switchover. That is the standard on every migration I run, not a paid extra.
From numbers, not from reassurance: Core Web Vitals before and after, index coverage and errors in Search Console, structured data passing the Google tests, and for AI visibility - a comparison of what the models said about the company before the changes and after them. I do not promise dates for ranking results, because that would be guessing.
Want to know what is blocking your visibility?
Describe the project in a few sentences - I come back personally with the scope, a fixed amount and a date.