Almost everyone in digital marketing shares one assumption: big cities are expensive, small ones are cheap. You pay a premium per click in İstanbul and get the same ad far cheaper in the provinces. Budgets are built on that assumption. So are agency proposals.
We believed it too. Our own Google Ads budget calculator had the logic hardcoded: give each industry a base cost per click, then multiply by a regional coefficient. İstanbul 1.25. Major cities 1.0. Regional markets like Aydın and Kuşadası 0.8. Smaller districts 0.65.
Then we measured it. The coefficient was not miscalibrated — it was backwards.
The finding: İstanbul is not the priciest healthcare market
We pulled the real median cost per click Google Ads reports for 7 industries across 12 Turkish provinces — 84 industry × province cells. For healthcare:
- Muğla — ₺114.7 (the most expensive healthcare market in the country)
- İstanbul — ₺102.1
- Ankara — ₺66.6
- Aydın — ₺57.8
- İzmir — ₺55.4
- Konya — ₺21.4
Muğla costs more than İstanbul. Aydın costs more than İzmir. Our calculator modelled Aydın at 0.8× İzmir; it actually measures 1.04×. Muğla we treated as a cheap "regional market" — it is the most expensive province in the set.
That is not a calibration error. It is a model pointing the wrong way. And we were using it in the industry we serve most, in our own region.
Why a thin market can cost more
Google Ads is an auction, and auction prices are set by supply as much as demand. A large city has high search volume, but it also has a large pool of impression inventory; bidders spread across it.
A thin market inverts that. Monthly search volume for our healthcare basket in Muğla is a fraction of İstanbul's — but the handful of clinics bidding for that small volume compete for all of the inventory there is. A few aggressive bidders in a narrow auction push the price up. And along the health-tourism corridor (Bodrum, Marmaris, Fethiye) those clinics carry very high value per patient, so they can afford to pay it.
In other words, "small city means cheap ads" breaks precisely in the industries where customer value is highest.
It does not hold across every industry
The limit of the finding matters. Provincial spread varies enormously by industry:
- Healthcare — 5.4× (₺21.4 Konya → ₺114.7 Muğla). The widest.
- Real estate — 4.4× (₺3.65 Konya → ₺16.04 Kocaeli)
- Restaurants and cafes — 4.4× (₺5.83 Kocaeli → ₺25.76 Adana)
- Education — 2.6×
- Furniture — 1.9×
- E-commerce — 1.6×
- Tourism — 1.4× (₺14.82 Denizli → ₺20.66 İstanbul)
In tourism the province barely matters: 1.4× is close to noise. The likely reason is that tourism competition is not local but national — often international. Someone searching for a hotel is not necessarily searching from the province they will visit.
So the right generalisation is not "small cities are expensive". It is this: province choice is the single most decisive budget variable in some industries and nearly irrelevant in others. Which one you are in is worth checking rather than assuming.
What this means for your budget
For a clinic targeting 25 qualified clicks per day, media spend alone:
- Konya: ~₺16,300/month
- İzmir: ~₺42,100/month
- Aydın: ~₺44,000/month
- Muğla: ~₺87,200/month
Same target, same industry, a spread of more than five times. If an agency tells you a regional campaign will be "more efficient", ask which industry and which province.
Method and limits
The source is Google Ads keyword planner data, pulled via DataForSEO. Values are 12-month averages; Google does not publish the current month. Each industry uses a customer-facing keyword basket — what a clinic, hotel or estate agency actually bids on. The cell value is the median of that basket, not its most expensive keyword.
Three limits worth stating plainly:
- A single pull. Every figure comes from one snapshot dated 2026-09-12. Unlike our Turkish word-order study, this data has not yet been replicated in an independent pull. It is too early to say whether the finding is structural or seasonal.
- Province level, not city. At city granularity Google returned no data for 4 of 5 healthcare keywords. The Muğla healthcare cell is built from just 3 of 12 keywords — the index publishes per-cell coverage so you can weigh each one.
- Exchange rate. Google reports cost per click in USD. TRY figures here use ₺48.60/USD. When the rate moves the TRY numbers move; the USD numbers do not.
Check the data yourself
All 84 cells, the methodology and a downloadable CSV are published free. The data is CC BY 4.0 — reuse it with attribution.
We are publishing this because we found our own tool was wrong and fixed it. If anyone else is planning budgets on the same assumption, they should at least get to decide against the numbers.
