Google Italy ranking factors 2026: what really counts? Our study on 600+ pages

Every year SemRush, Backlinko and Moz publish ranking factor studies based on English-language datasets. The problem is that Italian SERPs have their own dynamics: different domains, different search habits, and a competitive landscape dominated by local players that English-language datasets simply don't capture.

We decided to conduct the first correlational study on ranking factors specific to Google Italy, using real Italian keywords extracted from sites we manage or monitor, including wikiherbalist.com (herbalism and botany) and f-hack.com (cybersecurity and tech).

Methodological disclaimer: this is a pilot study with a limited dataset (601 URLs, 135 keywords, specific niches). Results are indicative, not definitive. Correlation does not imply causation: we measure statistical associations, not certain causal mechanisms. We plan to expand the dataset to 1,000+ keywords in the coming months.

Methodology

1. Keyword collection

We extracted opportunity keywords from our sites via Google Search Console: queries with an average position between 4 and 20 and at least 20 monthly impressions. The result: 135 Italian keywords across herbalism, cybersecurity and other niches, with real and verified search volume.

2. SERP scraping from google.it

For each keyword we extracted the top 10 organic results from google.it (it-IT locale) using Playwright with Chromium, the same engine as the Chrome browser. Total: 1,187 URLs collected from SERPs.

3. Page analysis

Each URL was visited and analyzed to extract 11 on-page and technical factors. Unreachable URLs were excluded. Final dataset: 601 pages analyzed.

4. Spearman correlation

We calculated the Spearman correlation coefficient between each factor and the SERP position. We chose Spearman over Pearson because the relationship between factors and position is non-linear: the difference between position 1 and 2 is not the same as between 9 and 10. Significance threshold: p < 0.05.

The results

The complete factor map

Negative values = correlation with better positions. * p<0.05 · ** p<0.01 · *** p<0.001

Only three factors are statistically significant in the Italian SERPs analyzed:

The most important factor: keyword match

Pages in positions 1–3 contain the keyword (or a variant) in the title, H1 and URL slug with significantly higher frequency than those in lower positions:

Pages in positions 1–3 include the target keyword in the title tag 28% of the time, versus 17% for those in positions 7–10.

This doesn't mean that simply inserting the keyword in these fields is enough to rank. It means that those already ranking well do it more frequently, and that keyword match in the main meta tags remains a relevant signal for Google Italy in 2026.

Practical case: wikiherbalist.com

For the keyword "saponins" (9,600 impressions/month, average position 9.6), the wikiherbalist page has the term in the slug and H1, but not in the title tag. Pages in positions 1–3 for this query have the term in the title 100% of the time. A title tag modification is the first optimization to make.

The data that debunks a myth: word count

One of the most common SEO tips is "write long content to rank better". Our data tells a different story, at least for the Italian SERPs we analyzed:

Pages in positions 1–3 average 1,949 words, fewer than those in positions 4–6 (3,383 words). The correlation is practically zero: r=-0.011, p=0.79.

How do you explain this paradox? By looking at who occupies the top positions: Wikipedia and Treccani dominate most Italian informational SERPs we analyzed. These sites don't need extremely long content to rank: domain authority does the work.

Sites that try to compensate for low authority with very long content (positions 4–6, averaging 3,383 words) can't outrank established authority sites. The takeaway isn't "write less", but that length alone is not a discriminating factor: who you are matters more than how many words you write.

Who dominates Italian SERPs: the competitor radar

Analyzing all the SERP results collected, a clear picture emerges of who holds the top positions:

The 10 domains with the most appearances in the top 10 positions across 135 Italian keywords.

Wikipedia appears in the top 10 positions for 83 out of 135 keywords (61%). Treccani for 67 (50%). These two sites are effectively the gatekeepers of Italian informational SERPs.

The right strategy for those competing in these niches is not to try to beat Wikipedia on generic queries. It's to target long-tail keywords with specific intent where the major authority sites are absent or unoptimized.

Practical case: f-hack.com

For keywords like "content security policy examples" (position 16, 130 impressions), neither Wikipedia nor Treccani appear in the SERP: the territory is contested by specialized technical sites. Here f-hack.com has real room to climb, by working on keyword match in the main tags and on the technical quality of the page.

Non-significant factors: what does NOT matter (in these SERPs)

Equally interesting is what is not correlated with positions:

• HTTPS (r=-0.032, p=0.43): 99.2% of the pages analyzed already use HTTPS. It has become a prerequisite, not a competitive advantage.
• Gulpease readability index (r=-0.034, p=0.42): Italian text readability shows no significant correlation with positions in this dataset.
• Freshness (r=+0.012, p=0.88): content update date is irrelevant in the niches analyzed. Authority sites maintain positions regardless of how recent their content is.
• .it TLD (r=+0.069, p=0.09): contrary to what is often recommended, the .it domain brings no systematic advantage in Italian SERPs. Wikipedia (.org) and many .com sites coexist at the top without meaningful differences.
• Images present (r=+0.073, p=0.072): borderline non-significant, but the direction is worth noting - pages with images tend to rank slightly lower. Likely a Wikipedia effect: the dominant authority sites are often plain text.
• Video present (r=-0.001, p=0.979): essentially zero correlation. Embedded videos show no association with positions in the SERPs analyzed.
• Title tag length (r=+0.004, p=0.916): the character length of the title tag is irrelevant. What matters is what it contains (keyword match), not how long it is.
• Structured data (schema.org): measured via JSON-LD detection, but returned 0 out of 601 pages. The main Italian authority sites (Wikipedia, Treccani) use microdata or RDFa instead of JSON-LD, which our parser doesn't detect. We will repeat the analysis with full detection in the v2 study.

Limitations and next steps

We are the first to acknowledge the limitations of this study:

• Limited dataset: 601 URLs and 135 keywords in specific niches. Correlations could differ in other sectors (e-commerce, local SEO, transactional queries).
• Unmeasured factors: domain authority (backlinks), user behavioral signals (CTR, dwell time), Core Web Vitals. These factors require data not accessible via public scraping.
• Correlation is not causation: the significant factors are associated with good positions, but are not necessarily their direct cause.

The v2 study is already underway. We have expanded the dataset to 900 new keywords generated via Google Autocomplete across 15 diverse niches - food, health, finance, tech, travel, home, work, education, legal, automotive and others - bringing the total to over 1,000 Italian keywords. The new data will be integrated into this post once the analysis is complete.

The full study will also introduce new factors, some already measurable via the insights.perseodesign.com crawler:

• Core Web Vitals (LCP, CLS, INP): Google uses these performance signals as a ranking factor. We will measure the actual values of each competitor page via Lighthouse.
• Domain authority: via the free OpenPageRank API we will obtain a backlink profile proxy - likely the most influential factor this pilot study could not measure.
• Internal links to the page: the number of internal links pointing to a page reflects the weight the site itself assigns to it.
• Heading structure: H2/H3 count, keyword presence in H2 tags, heading hierarchy depth.
• Image alt text: whether the target keyword appears in the alt attribute of the main images.
• Page depth: the number of URL levels as a proxy for distance from the homepage and perceived importance in the site structure.
• Full structured data: detection of JSON-LD, microdata and RDFa to properly cover Italian authority sites that don't use JSON-LD.
• E-E-A-T signals: presence of author, publication date and editorial byline - increasingly relevant with the rise of AI Overviews.

Suggestions

Based on the data collected, the actions with the best effort-to-impact ratio for an Italian website are:

1. Optimize title, H1 and URL with the target keyword. It's the factor most correlated with good positions. Before anything else, verify that every page has the keyword (or a natural variant) in these three elements.
2. Choose keywords where authority sites aren't competing. Competing with Wikipedia on generic queries is a losing bet from the start. Identify the long-tail queries where the field is open.
3. Don't obsess over content length. Write what's needed to answer the query. Length alone doesn't shift positions in the Italian SERPs we analyzed.
4. Monitor your actual organic competitors. Who really appears in your SERPs? Often they're not the brands you perceive as competitors, but niche sites you may not even know.

Full methodology: scraping with Playwright/Chromium from google.it (it-IT locale), HTML analysis with BeautifulSoup/Python, Spearman correlation with scipy. Dataset: June 2026.

Want to analyze your site's SEO opportunities? Try insights.perseodesign.com: technical analysis, opportunity keywords from GSC and comparison with competitors in your SERPs.