Hospitality · Algiers · English, French, Arabic
Hydra Hotels
Three hotels in Algiers, one domain, three languages and no website to start from. Sixteen months after launch the site has taken 12,800 clicks from 605,000 impressions and ranks for the full thousand queries Search Console will report.
This is the whole picture, including the parts that are less flattering than the headline. Every figure below comes from Search Console and nothing has been modelled or rounded up.

Source: Google Search Console, web search, 16 month window to October 2026. Average CTR across all queries 2.1%.
Live site: hotelhydra.dz
The starting point
Hydra Hotels operates three properties in Algiers: Sidi Yahia, Ouled Fayet and Kouba. The group had rooms, conference halls and a restaurant, and no website.
That makes the baseline unusually clean. There was no legacy site to migrate, no old rankings to protect and no previous agency's work to unpick. It also means there is no before-and-after percentage to quote, because the before was zero. Everything below is what a domain built from nothing accumulated in sixteen months.
The complication was Algiers itself. Algeria is francophone with Arabic as the official language, and a hotel group serving business travel also needs English. One site, three audiences, three scripts, and a market where the same hotel is searched for in at least two spellings.
What was built
- A website from scratch, with the booking path as the primary journey rather than an afterthought bolted onto a brochure.
- Three full language versions, English, French and Arabic, each at its own path with hreflang declared so Google serves the right one rather than guessing.
- A page per property, because Sidi Yahia, Ouled Fayet and Kouba are three different locations competing for three different sets of local searches.
- Right-to-left support for Arabic, which is a layout problem rather than a translation problem and is where most multilingual builds quietly fail.
- Local SEO foundations per property: consistent name, address and phone, and structured data describing a hotel group rather than one generic business.

The results
Sixteen months of web search data, to October 2026.
| Metric | Figure |
|---|---|
| Clicks | 12,800 |
| Impressions | 605,000 |
| Average CTR | 2.1% |
| Average position | 7.3 |
| Queries reported | 1,000 (the Search Console maximum) |
| Monthly average | roughly 800 clicks and 37,800 impressions |
The ten queries doing the heavy lifting

| Query | Clicks | Impressions | CTR |
|---|---|---|---|
| hotel hydra | 1,636 | 24,658 | 6.6% |
| hotel hydra alger | 1,635 | 49,688 | 3.3% |
| hotel hydra ouled fayet | 1,088 | 14,329 | 7.6% |
| hydra hotel | 438 | 7,258 | 6.0% |
| hôtel hydra alger | 312 | 3,333 | 9.4% |
| hôtel hydra sidi yahia | 252 | 3,296 | 7.6% |
| hydra alger quartier chic | 204 | 1,114 | 18.3% |
| hydra hotel alger | 188 | 1,920 | 9.8% |
| hotel hydra contact | 174 | 899 | 19.4% |
| hôtel hydra | 169 | 3,413 | 5.0% |
What the data actually shows
Ten queries, half the traffic
Those ten queries account for 6,096 clicks, about 48% of the total, from 109,908 impressions, about 18% of the total. The head converts at three to four times the rate of the tail. That gap is the entire argument for building a site around the handful of searches that actually produce business rather than chasing query count.
The accent is a separate market
Look at rows two and five. hotel hydra alger took 1,635 clicks from 49,688 impressions. hôtel hydra alger, the correctly accented French spelling, took 312 from 3,333. Google treats them as different queries and people type both, often depending on whether their keyboard makes the circumflex easy.
Miss the accented variants and you lose a real slice of a francophone market. The unaccented form carries far more volume but the accented one converts better, at 9.4% against 3.3%, which suggests the people who bother with the circumflex are further along in deciding.
Location terms rank independently
hotel hydra ouled fayet took 1,088 clicks at 7.6% CTR, and hôtel hydra sidi yahia another 252. Two of the three properties pull meaningful traffic on their own location names. That only happens because each property has its own page. On a single combined locations page those searches would have landed somewhere generic and converted far worse.
The highest CTR is someone trying to call
hotel hydra contact has the best click-through rate in the table at 19.4%, and hydra alger quartier chic is close behind at 18.3%. Small volumes, but these are people who have already decided and are looking for a phone number or confirming the neighbourhood. In hospitality that is the query closest to a booking, and it is worth more than its click count suggests.
What this does not prove
Search Console measures search, not business. It can tell you how many people arrived and what they typed. It cannot tell you how many of them stayed the night.
- No revenue or booking figures are published here. That data belongs to the client, not to me, and I am not going to estimate it to make a better headline.
- There is no before-and-after comparison, because there was no before. A site built from scratch starts at zero, which makes the growth real but the percentage meaningless.
- An average position of 7.3 is an average across a thousand queries, including a long tail that sits well down page one and beyond. It is not a claim that the site ranks seventh for anything in particular.
- Brand searches are included. A good share of
hotel hydratraffic is people who already know the hotel. Building the brand's search presence is part of the job, but it is not the same as winning a customer who had never heard of them.
What transfers to a UK business
Algiers is not London and a hotel group is not a dental practice. What transfers is the structure of the problem, and it is the same one most multi-location businesses have.
- A page per location, written individually. Three properties produced three independent streams of location-name traffic. Branches sharing one page share one ranking, which in practice means they compete with each other.
- Spelling and phrasing variants are separate markets. The French accent problem here is the same shape as the UK one, where people search both the trade name and the plain-English description of the job.
- Navigational queries are the warmest traffic you have. Whatever the local equivalent of
hotel hydra contactis for your business, it converts better than anything else and is usually ignored. - Entity consistency across versions. Three languages here, but the same discipline applies to any business whose details appear in more than one place. See entity building.
If you run more than one location, or more than one language, the first useful step is finding out where your coverage actually stops. That is what a free audit measures.
Straight answers
Questions
How long did it take to reach these numbers?
The figures cover a sixteen month window to October 2026, starting from a domain with no website on it. Impressions build before clicks do, and the pattern here was the usual one: a slow first quarter while pages were indexed and assessed, then compounding growth. There is no shortcut to the compounding part.
Why build three languages rather than one and translate later?
Because retrofitting languages is considerably more expensive than designing for them. URL structure, hreflang, right-to-left layout and the entity consistency between versions all get decided at build time. Bolting Arabic onto a site that was never laid out for right-to-left usually means rebuilding the front end.
Does any of this work for a single-location business?
Most of it. The multi-location and multilingual parts do not apply, but the head-versus-tail finding, the spelling variants and the value of navigational searches all hold for a single site. The free audit is where you find out which of them is costing you.
Can you share booking or revenue numbers?
No. That data is the client's and publishing it is their call rather than mine. What I will publish is the search data, in full, including the average position and the caveats, so you can judge the work rather than take a headline on trust.
