Local business websites 2026: The technical state of 847 websites (Onur Freelance research)

The four technical basics: Only 40% of websites pass all four checks

All four of HTTPS, a mobile viewport tag, a meta description, and a single H1 heading are present on 40% of the 847 websites. HTTPS and the mobile tag do not pull this combined measure down; the main losses concern meta descriptions and especially H1.

Websites not opening with HTTPS are a minority at 2.8%, as are those without a mobile viewport tag at 3.1%. Google states that it uses the smartphone-crawled version of a website for indexing and ranking and recommends responsive design for its ease of maintenance in its mobile-first indexing documentation . A viewport tag is a prerequisite for responsive design, but alone does not prove that the page displays correctly on a phone.

Meta descriptions are missing on 20.3% of websites. According to Google’s snippet documentation, search result descriptions are mostly generated from page content; meta descriptions are used when they describe the page more accurately than the content. The same document recommends starting with priority addresses such as the homepage if there is no time to write descriptions for every page; that is exactly the page this research examines.

Contact: WhatsApp links, clickable phone numbers, and websites with neither

76.7% of websites have a clickable phone link, 49.7% have a numbered WhatsApp link, and 19.8% have neither. In other words, one in five websites offers visitors on a phone no one-tap route to calling or messaging.

“Neither” does not mean no number is written on the website. A number may appear as plain text or an image; visitors must enter it manually. Because measurement counts only links beginning with tel: and WhatsApp addresses containing a number, a WhatsApp message link without a number also appears as “absent.”

These two links are among the least labor-intensive improvements on the list. Requiring only a few lines of changes to implement, the links shorten the contact path alongside free tools such as the Google Business Profile review link .

Measurement code: Half the websites show no trace of Google Analytics

49.1% of homepage sources contain traces of gtag, Google Tag Manager, or Google Analytics; the remaining 50.9% show none of these three. On a website without measurement code, it is impossible to find out retrospectively which search, advertisement, or page generated phone calls.

The limitation of this percentage is clear: websites using Yandex Metrica, Hotjar, or server-side measurement may have been counted as “absent.” Websites not loading tags at all until cookie consent also fall into this group if they leave no trace in the source. The percentage should therefore be read as “websites with a Google measurement tag visible in homepage source.”

For businesses with advertising budgets, the gap costs more: read together with our Google Ads click costs research , which shows click costs by industry, it also means the return on this spending cannot be tracked on an unmeasured website.

H1 heading structure: Fewer than half use a single H1

47.8% of the 847 websites use a single homepage H1; 35.2% have none, and 17% have multiple H1s. This item pulls the four-basics measure down the most.

Google’s title links documentationstates that it may use the main visual page title and heading elements such as h1 alongside the title element when generating search result titles. The document notes that multiple large headings with equal visual weight make the main heading unclear and recommends clearly distinguishing the main heading from other text in the page’s first h1.

Two situations common in ready-made themes may explain these percentages:

• Coding the logo or slider headings as h1; outputting the same heading twice for mobile and desktop.
• Writing the homepage’s large text as an ordinary text block or h2 instead of h1; the area looks like a heading but is not a heading in code.

A page with no H1 or multiple H1s may continue appearing on Google; this is an organizational issue clarifying what the page describes rather than a penalty matter.

Structured data: JSON-LD blocks on half the websites

49.6% of websites contain a JSON-LD structured data block on their homepage. Block types and validity were not checked; generic markup automatically added by WordPress SEO plugins also counted as “present.”

Google’s introduction to structured datastates that JSON-LD, microdata, and RDFa are supported and recommends JSON-LD because it is easy to implement. Since this research searched only for JSON-LD, websites using microdata may have remained in the “absent” group. For local businesses, markup consistently describing the business name, address, phone number, and opening hours is generally the first step.

Recency, speed, and infrastructure: Copyright year, response time, WordPress

The footer copyright year is 2023 or earlier on 21.7% of websites and 2025 or later on 37.4%; on the remaining 40.9%, the year is 2024 or was not found in the source. Copyright year is an indirect sign of recency: an old year does not prove abandonment, merely offering a weak clue that the website has not been reviewed for some time.

Homepage response time

Median time for homepage HTML to arrive from the server was 577 milliseconds, and median HTML size was 82 KB. On 19.2% of websites, this time exceeded 1.5 seconds. This is not a full page load measurement; images, scripts, and fonts were excluded, and each website was measured once from one location. Momentary server load may affect results; this percentage is not a judgment on individual websites.

WordPress share

29.8% of websites contained wp-content in their source, indicating WordPress use. The remaining 70.2% may be ready-made website platforms, custom software, or WordPress installations with hidden traces; this distinction was not made.

Is there a difference between websites in the top 10 and positions 11–30?

Differences are small for most metrics; a clear distinction appears only in whether the searched city’s name occurs on the website. The searched city name never appears on the homepage of 2.5% of the 565 websites in the top 10 and 9.6% of the 282 websites in results 11–30.

This is a relationship, not cause and effect. It cannot be said that mentioning the city name on the page produces a higher ranking; most of the top 10 group comes from Yandex local results, while most of the 11–30 group comes from Google, and local results naturally highlight businesses in that city. Websites in positions 11–30 without the city name may also be companies from another province appearing in that search.

Other differences are smaller. The differences in structured data (52.2% versus 44.3%) and multiple H1s (14.7% versus 21.6%) are 7–8 percentage points; single H1 is 49.4% versus 44.7%, meta descriptions 80.2% versus 78.7%, and measurement code 49.6% versus 48.2%. The 11–30 group leads for WhatsApp links (47.4% versus 54.3%) and HTTPS (96.8% versus 97.9%). This data cannot isolate which signal affects ranking.

Business website shortcomings by industry

Across 21 industries with at least 15 websites, the most divergent metrics were single H1, measurement code, and WhatsApp links. In industries with samples of 17–25 websites, one website shifts percentages by 4–6 points; read the table as trends rather than rankings.

73.9% of plastic injection molding websites have no homepage H1, and their meta description rate is the lowest at 56.5%. Elevator websites have the highest measurement code rate (66.7%), but also the highest share with copyright years of 2023 or earlier, at 48.5%. The searched city name does not occur on 23.1% of vehicle inspection and 30.4% of shelving systems homepages; multi-branch companies or companies serving from outside the province appearing in these searches may explain this.

A brief note by city

With provincial samples ranging from 21 to 77 websites, differences should be interpreted cautiously. Measurement code is highest in İzmir at 65.2% and lowest in Sakarya at 31.1%. Bursa has the lowest provincial rates for WhatsApp links at 26.9% and clickable phone links at 58.2%; Antalya has the highest single-H1 rate at 65.6%, and Samsun the lowest at 29%. Meta descriptions appear on 89.8% in Adana and 68.6% in Mersin.

How can a business perform these checks on its own website?

Most research checks can be performed in a few minutes by opening the homepage in a browser, viewing page source (Ctrl+U on Windows), and searching within it. The sequence can be:

1. HTTPS: Does the address open with https:// in the address bar, and does entering http:// redirect to https?
2. Mobile tag and display: Does the source contain name=”viewport”; can the website be read on a phone without horizontal scrolling?
3. Meta description: Does the source contain name=”description”, and does its content describe what the business does and which city it serves?
4. H1: How many results does searching “<h1” in the source return? Ideally one; that heading should describe the service and, if possible, the city.
5. Phone and WhatsApp: Does tapping the number start a call; is the WhatsApp button a numbered wa.me link?
6. Measurement: Does “gtag” or “googletagmanager” appear in the source; does the Analytics account show data for recent days?
7. Structured data: Does the source contain “application/ld+json”; does Google’s Rich Results Test show errors?
8. Recency and city: Is the footer year current; do the province and district served appear in the page text?

Subpages, page speed, and indexing status require separate examination. To preview title and description drafts before publication, tools on the free tools page can be used; more comprehensive technical crawls are performed within the technical SEO service .

To plan which month to address shortcomings, industry seasonality data shows the busy season. Onur Freelance is not an official Google or Yandex support channel; the research is based on an independent crawl and may be updated in a subsequent round.