How to Choose Keywords for a Client from Sibiu or Bucharest

A dentist gave us their keyword list. In first place: "dental implant". A term with national competition, sought after by people doing research, sometimes from another corner of the country.

What actually brought in patients, when we looked at the data a few weeks later, were less impressive searches: "emergency dental Sunday", "wisdom tooth extraction price", the name of the neighbourhood next to "dentist".

Keyword research for a local business is less about volume and more about understanding what stage the person searching is at.

Where we start the list

From three sources, in this order.

What you tell us: how your clients name your services, what they ask you on the phone, what words *they* use, not you. The difference between a specialist term and the word used by the client is often bigger than you think.

What's already in your data: the search terms report from Google Ads, if you have an account, and search performance reports, if the site has history. These are the best sources, because they're real.

Tools, last. They give us approximate volumes and variations we hadn't thought of. Volumes for small towns are imprecise anyway, so we treat them as an indication of order of magnitude.

How we group

Not by volume. By what the person wants.

Someone doing research ("what is a dental implant") gets an article. Someone comparing ("dental implant or bridge") gets a page that honestly compares, including cases where the cheaper option is suitable. Someone who's ready ("dentist Sibiu emergency") gets the service page, with phone number and schedule clearly visible.

Each group goes to one page, one page per group. When you see two pages on the same subject, they usually compete with each other.

What we leave out

Generic, high-volume terms that you can't win in the next year and that anyway bring people from other parts of the country. They look good in a report. They don't produce, and how quickly you can rank for them depends on the niche and the age of the site.

We don't treat variants with and without diacritics separately, because the search engine understands them. We don't build duplicate pages for them.

And something we did wrong: we once built a plan with 200 terms for a company with two employees. Nobody had time to produce that content. Now the plan has 15-20 ordered topics, and we expand it when we see the pace being kept. We rethought our planning process exactly because of this.

The difference between cities

A client from Sibiu and one from Bucharest don't have the same problem. In the big city, competition for service terms is high, but there's enough volume for very specific, neighbourhood-based searches. In the smaller city, volume is low for everything, but you can cover almost everything if you have a few months of patience.

This also changes expectations. I can't promise someone from Bucharest the same speed of results as someone from Sibiu, on the same budget.

If you want, we can look at what searches are already bringing people to your site and tell you what's missing.

What we do with pages, based on what people search for

Grouping words isn't a spreadsheet exercise, but it ends up in a spreadsheet, because otherwise you can't decide how many pages to write.

The last two are often confused. We once wrote a long article for a search that was, in fact, "I want a company now". It brought visits and no enquiries. We turned it into a service page and that changed.

Tools, and how much we rely on them

We look at what volume tools show, but we don't take the figures as gospel. For small towns, reported volumes are approximate and sometimes simply wrong. What we do additionally: we ask the client how people ask for the service on the phone. From there comes a list you won't find in any tool.

A plumber told us that people don't search for "plumber", they search for "I don't have hot water". It didn't appear anywhere on our initial list.

What we leave aside in the first few months

Very competitive, general searches that we can't rank for quickly. We don't ignore them completely, but we don't build the beginning on them. We leave them for when the site already has pages that are ranking.

How long it takes to see results

It depends on how old the domain is and the competition. For new domains, the first signs appear on long and specific searches. I can't give you an exact date, and we don't like it when someone does.

If you want to see what people in your niche are actually searching for, we'll make you a grouped list and show you what pages might be missing from your site.