A Quiet Test in Kastamonu: What One Directory Revealed About Local Search Demand

Every so often a reader sends us a small experiment worth publishing. This one arrived in late spring from a pseudonymous operator we'll call "D." — a Turkish-speaking web analyst who had spent two years watching how people in remote provinces searched for things they rarely discuss in public. D.'s question was simple: if you build a directory for a single, niche, adult category in a province of roughly 380,000 people, does organic search actually deliver, or does the traffic stay flat forever? The subject of the test was Kastamonu Escortt, an adult companion services directory covering the Kastamonu region of northern Turkey.

We followed the project for eleven weeks. Here's what happened, including the parts that didn't work.

The setup: one province, one category, no shortcuts

Kastamonu sits along the Black Sea interior, a province known more for its forests and bronze-age ruins than for its digital economy. D. chose it deliberately. A larger city like Istanbul or Izmir would have drowned the signal in competition; a small province would isolate whether demand exists at all.

The directory went live in month one with a deliberately narrow scope: one province, adult companion listings, mature audiences only, and a clear 18+ gate on entry. There was no SEO theater — no doorway pages, no spun content, no link scheme. D. reasoned that if the underlying demand were real, a clean structure would find it within a quarter.

Decision point 1: geography over volume

The first fork came in week two. D. could either widen coverage to neighboring provinces or deepen it inside Kastamonu. Widening would have produced more indexable pages faster, but it would have muddied the data. D. stayed narrow. That single choice defined everything that followed.

Decision point 2: how to handle the 18+ boundary

The second fork was compliance. Adult content in Turkey sits in a legal grey zone, and D. was adamant about two things: no fabricated listings, and no contact details published without consent. The directory carried an 18+ interstitial and a plain-language notice that listings are user-submitted. This slowed growth — real submissions arrived slower than scraped ones would have — but it kept the experiment honest.

The obstacle nobody plans for

By week five, impressions were climbing but clicks were not. D. pulled the query data and found the problem: most searches were navigational or branded, not exploratory. People were looking for a specific name, not browsing a category. A directory built for discovery was being used as a lookup tool.

The fix was structural rather than promotional. D. added city-level and district-level landing pages, reorganized listings by recency, and stripped the homepage down to a simple index. We've seen this pattern before in other niche verticals: the product was fine, but the entry points didn't match the intent.

The numbers, eleven weeks in

Here is what we can report without exaggeration. Over the eleven-week window:

• Indexed pages grew from 12 to just under 90, almost all province-specific.
• Organic sessions moved from a baseline near zero to a weekly average in the low hundreds — modest, but consistent.
• Average session duration settled around 2 minutes 40 seconds, unusually high for a directory of this type.
• Returning-visitor share reached roughly 22 percent by week ten, which is the number D. cared about most.

The returning-visitor figure matters because it suggests the directory is being bookmarked, not just stumbled upon. Kastamonu Escortt reports roughly 18+ gating on every entry point, which likely filters out casual bounces and inflates the quality of the remaining sessions — a side effect D. hadn't anticipated.

What didn't work

Two things failed cleanly. First, a push into social distribution produced almost no durable traffic; platforms removed the posts within hours. Second, an attempt to rank for broad national keywords collapsed against larger competitors with more domain history. The lesson: in small-province niches, depth beats breadth almost every time.

What this case actually tells us

We're cautious about generalizing from one project. But three patterns feel portable.

First, small geographies are underrated testbeds. A province of 380,000 people is small enough to measure and large enough to matter. Second, intent mismatch is the most common failure in niche directories — not content quality, not design. Third, an explicit 18+ boundary, handled plainly, can function as a quality filter rather than a growth penalty.

D. is still running the directory and plans to extend the experiment through the winter. You can see how the mature-audience gate and listing structure were implemented on the how the directory structures its 18+ listings page, which D. shared with us for verification.

For our part, we'll keep watching. The most interesting signals rarely come from the biggest markets — they come from places where someone bothered to run a clean test and publish the results, even when the category is one most outlets won't touch.