The difference from Moldova is brutal: in Romania, "website development" or "lawyer Bucharest" are contested by firms with real marketing budgets.
You do not win there starting from zero. What can be won — and what actually brings clients — are the narrow searches with clear intent that nobody covers because they look too small.
Where the space is
Specific intents. "Website for a dental practice", "Passat clutch replacement price", "post-renovation cleaning". Each has low volume and almost no competition, and someone searching that way is a step from buying.
Service plus city, beyond Bucharest. Cluj, Timișoara, Iași, Brașov, Constanța. The same searches, far less competition.
The questions asked before buying. What it costs, how long it takes, which documents are needed, what happens if. They do not bring large traffic, but they bring exactly the people about to decide.
What has to be fixed first
None of the above matters if the pages are not readable.
Rendering. A JavaScript site serving empty containers appears nowhere, however well the copy is written.
Canonicals, pagination, filters. Large shops generate thousands of URLs from filters. Without correct canonicals the index fills with variants of the same page and no version ranks.
Speed on a phone, on mobile data. In a competitive market it is a real differentiator, not a checkbox.
Product structured data. Without it the shop loses rich results — and the clicks that come with them.
Visibility in AI answers
More decisions now start with a question to an assistant rather than a search. Those systems do not answer from memory; they answer from what they find at the moment of answering: ChatGPT searches through the Bing index, Gemini through Google.
The practical conclusion is the same: a page a crawler cannot read enters none of those answers. There is no separate "SEO for AI" — there is the same work, verified additionally against GPTBot, ClaudeBot and PerplexityBot.
What we do not promise
Positions, because they cannot be guaranteed. Results within weeks on commercial terms — the technical part yes, authority no. And we do not buy links: they are an explicit violation with a risk of manual action that takes months to recover from.
How we verify
Against real crawlers — Googlebot, Bingbot, GPTBot, OAI-SearchBot, ClaudeBot, PerplexityBot. The tool is public: crawlview.
Questions
Can we compete on general terms?
Not from zero and not in the first year. The strategy that works is to accumulate positions across dozens of specific searches, and the resulting authority makes general terms possible later.
Why do you not buy links?
Because they are an explicit violation and because a manual action costs more than they bring. We wrote here about what the ecosystem selling them looks like.
How long?
Technical fixes show within weeks. Positions on specific searches, within months. On competitive general terms, years or never, depending on resources.
What do we actually receive?
A document with the technical problems and their order, a map of the intents you can realistically cover, and verification against real crawlers before and after.