İçeriğe geç
Flatinium

Open data · October 2026

Business websites in Turkey’s three largest cities — what they miss

We scanned 937 business websites in Ankara, Istanbul and İzmir with identical code and criteria. The cities look remarkably alike: about half of the sites have no click-to-call phone link and two thirds do not describe the business in structured data. The only clear difference between cities is WhatsApp.

By Flatinium, an SEO and AI search optimization agency in Ankara · Türkçe sürüm

Key findings

  • Half of the sites show a phone number that cannot be tapped to call (Ankara 50%, Istanbul 52%, İzmir 50%).
  • About two thirds have no LocalBusiness or Organization structured data (63%, 62%, 69%).
  • WhatsApp is the one real city difference: no WhatsApp link on 76% of Ankara sites, 65% in Istanbul, 68% in İzmir. In food and drink, 89% of Ankara sites lack it.
  • AI crawlers are rarely blocked (3% or less in every city); 16–18% of sites publish an llms.txt file.
  • Most other city differences are within the margin of error (about ±5–7 points), so we do not present them as differences.
Chart: share of business websites missing WhatsApp link, business schema, click-to-call phone, H1, sitemap and meta description in Ankara, Istanbul and İzmir (labels in Turkish).
Chart labels are in Turkish. Free to reuse under CC BY 4.0 with attribution.

All criteria by city

Share of sitesAnkara (n=313)Istanbul (n=416)İzmir (n=208)
Opens over HTTPS95%94%96%
Has a mobile viewport tag97%97%96%
No meta description35%31%38%
No H1 heading42%47%44%
No business structured data (LocalBusiness/Organization)63%62%69%
No Open Graph preview41%42%41%
Phone number is not click-to-call50%52%50%
No WhatsApp link76%65%68%
No XML sitemap37%36%38%
Has an llms.txt file16%18%16%
Blocks at least one AI crawler1%2%3%
Homepage responds in over 3 seconds10%7%7%

Method

Business lists come from OpenStreetMap (businesses with a website tag within each province). Social media and marketplace URLs, public institutions, political parties, embassies, hospitals, associations, religious and educational institutions were excluded. Ankara (439 domains) and İzmir (293) were scanned in full; in Istanbul, 600 of 2,502 eligible domains were drawn with a fixed random seed. Only sites returning HTTP 200 with more than 30 words on the homepage were analysed. Scans ran on 1–2 October 2026 from Turkey.

Limitations

  • The sample is not random: businesses listed in OpenStreetMap lean towards established ones.
  • OpenStreetMap coverage differs by city; only the homepage was measured.
  • Response time is server response, not full page load.
  • Site names are not published; only aggregate rates.

Data and citation

Summary data (CSV), scanning code and methodology are open on GitHub (data CC BY 4.0, code MIT). Download the CSV directly. Please cite as: “Flatinium, Ankara–Istanbul–İzmir business websites scan, October 2026” with a link to this page. Business list © OpenStreetMap contributors (ODbL). Press contact: info@flatinium.com.