SEO explained for small businesses in Greece

SEO is not a trick you apply to a finished website. It is the sum of four things Google does with your page, in order: crawl it, index it, rank it and serve it. Every problem you will ever have with search lives in one of those four steps. Fix the step that is broken and stop guessing about the rest.

Step	What Google does	What usually goes wrong	How you check it
Crawl	Finds the URL and downloads the HTML	A noindex left on, a blocked path in robots.txt, a site that needs JavaScript to show its content	Search Console → URL Inspection → live test
Index	Stores the page in its database	Duplicate or thin pages, conflicting canonicals, an empty category page	Page indexing report
Rank	Scores the page against a query	Wrong intent, no real answer, nothing that makes you different	Impressions vs clicks per query
Serve	Shows the best result for that person	Title rewritten into nonsense, slow page, broken mobile layout	Title link and Core Web Vitals reports

Crawl: Google has to be able to read the page

Googlebot follows links and reads HTML. If your navigation is built by a script that never runs for Google, or a template ships <meta name="robots" content="noindex"> on every page, nothing else you do matters.

The check is the URL Inspection tool in Search Console. Paste the URL, run the live test, and read the rendered HTML, not the source. That is what Google sees. Google’s own overview of how Search works is worth twenty minutes once.

Index: the step most Greek sites fail

A page can be crawlable and still not indexed. Google decides a page is not worth storing when it looks like five other pages on the same site, when it has almost no text, or when it is a filter page that produces a thousand near-identical URLs.

An e-shop with 400 products, six filters and three sort orders can generate tens of thousands of URLs from 400 pages of content. Google crawls a fraction of them and indexes fewer. The fix is not more pages. It is canonical tags, a clean sitemap and fewer ways to reach the same product. The Page indexing report lists every excluded URL with a reason, and each reason has exactly one correct response.

If a page you expect to see is not in the index, read Why my page isn’t on Google before you change anything.

Rank: relevance, then everything else

Once a page is indexed, Google compares it to a query. Intent comes first: someone searching “ενοικίαση αυτοκινήτου Ηράκλειο” wants a rental form and a price, not a history of the automobile. A page that answers a different question can be perfectly optimised and still never appear.

After intent, the signals that move a small site are unglamorous. The words on the page match how people search, the page has one clear subject, other pages on the site link to it, and it loads quickly on a phone with a weak connection. Google publishes the canonical URL guidance because duplicate pages are the most common self-inflicted wound we see.

Speed is measured, not felt. Core Web Vitals is the score Google uses, and a page that shifts while it loads fails it even if the design looks calm on a fast laptop.

Serve: the result is not only your page

On a phone, Google shows a title it may have rewritten, a description it may have invented from your text, and a list of other results. If your title is 90 characters, the last words are gone before anyone reads them. That is why we treat the title and description as text to measure, not text to fill.

Why Greek sites lose before they start

Three local problems come up in almost every audit we run.

First, tonos. Google matches Greek queries with or without accents, so “ενοικίαση” and “ενοικιαση” usually return the same results. Your analytics do not: the two arrive as separate rows, and a keyword that looks weak may be split in half. Search Console shows this clearly once you look for it.

Second, Greeklish. A share of Greek searches still arrive in Latin letters, especially from people abroad or in a hurry. A site that only exists in Greek script misses those queries entirely. We built a free variants tool that writes one phrase in every spelling, because we needed it ourselves.

Third, directories. A hotel in Chania does not only compete with other hotels. It competes with Booking.com, and a product page competes with Skroutz. You will not outrank a marketplace for a generic term, and you should not try. Win the specific queries (the village, the service, the model, the problem) where a marketplace has nothing to say.

What this costs

SEO is not priced per hour here. A landing page starts at €290 and goes live in about seven working days; a business website starts at €490 and about fourteen working days; an e-shop starts at €890 and about twenty-five working days; a booking system starts at €1490 and about thirty-five working days. Extra pages are €149 for six to ten pages and €349 for eleven to twenty-five, a second language is €129, and all prices exclude VAT. The quote calculator gives the full number in two minutes, without an email address.

When a page does not index or does not rank, the diagnosis is the work. That is also why we ask for the URL first: Kourites Car Rental in Stalis, Crete had a site that loaded fine and a booking flow that worked, and the gain came from fixing what Google could not read, not from adding pages.

Where to start this week

Open Search Console. Check one page you care about in URL Inspection, and read whether it is indexed. Then look at the queries where you have impressions and no clicks. Those are pages Google already considers relevant, and they are the cheapest wins you have.

The next page to read is keyword research in Greek, which turns those queries into pages. If you want a price for the work instead, the quote takes two minutes.