What 291 training-center websites teach us about technical SEO and local rankings

We asked a simple question and turned it into an experiment. If we crawl an entire local-SEO sector, that of training centers and gyms across Spain, and measure its technical health variable by variable, what separates the winners from those who barely show up on the results page?

The short answer is uncomfortable: in local SEO almost nobody does the technical part well, and that is exactly where the opportunity lies.

The long answer is this article. We audited CMS, security headers, structured data, robots.txt, sitemap.xml, broken links, accessibility and local signals across 291 real websites, and cross-referenced them with their search position. What we found calls several myths into question (including one you probably defend yourself) and confirms a truth the industry repeats but few apply. A solid technical foundation is not an "extra" in local SEO. It is the ground everything else stands on.

How we read the data (and why it matters)

Before the conclusions, a piece of honesty rarely seen in SEO studies. Our sample has selection bias. These are websites that already rank for their local queries, so their best position is concentrated in a narrow range: almost all sit between position 1 and 5. When every subject in a study has already passed the exam, it is hard for a single variable to explain who scores a 9 and who scores a 6.

This has two implications that run through the whole article.

First: correlations with position will be weak by design. Not because the technical part doesn't matter, but because we are comparing survivors against survivors. The real effect of technical SEO shows up when you compare those who appear against those who never appear, and the latter are not in the sample.

Second: when a variable does show an advantage despite that bias, the signal is strong. Spoiler: exactly one pulls it off, cleanly.

With that in mind, on to the six questions.

1. Which CMS dominates the sector? WordPress, and not by a little

Of the 291 domains analyzed, 73 % run on WordPress (213 sites). Next comes a 19 % of custom or unidentified solutions (54 sites), and far behind Wix (7), Next.js (5), Drupal (2) and isolated cases of PrestaShop, Framer, Webflow and Squarespace. Within WordPress, Elementor is the recurring page builder.

This is no surprise. Globally, WordPress powers 43 % of the entire web and 61 % of all CMS-built sites, more than all its competitors combined (W3Techs / Automattic). In this local niche the concentration is similar, and arguably even more pronounced once you add the weight of WordPress templates plus visual builders.

Why does it matter for technical SEO? Because the ceiling and the floor of the sector are set by the WordPress ecosystem: the same templates, the same SEO plugins, the same default decisions. If the template produces mediocre structured data or leaves security headers unconfigured, that flaw multiplies across the whole sector. The good news is the mirror image. Careful technical configuration on WordPress is perfectly achievable, and it lets you stand out without reinventing anything.

2. Are Kit Digital websites worse? They tie on the bad and lose on what matters

Kit Digital has been the largest SME digitalization program in Spain's recent history: more than 860,000 grants awarded and €3.067 billion mobilized with Next Generation EU funds, with coverage reaching 92 % of municipalities (Acelera pyme / Red.es). We detected program-funded sites by their characteristic visual fingerprint: 32 of the 291.

The hypothesis was reasonable: subsidized website equals generic template equals worse technical SEO. The data qualifies it, and the result is more interesting than expected.

• Practically identical position: 2.88 on average for Kit Digital sites versus 2.80 for the rest. A technical tie.
• Same security debt (if anything slightly lower): 4.16 missing headers of 5 versus 4.28 for the rest.
• More raw structured data: 93.8 % of Kit Digital sites have some schema, versus 77.2 % of the rest, with more types on average (8.7 versus 7.2).
• And yet, less than half the `LocalBusiness` schema: only 9.4 % of Kit Digital sites implement it, versus 21.6 % of the rest.

There's the paradox. Kit Digital sites are not worse on the generic stuff (they actually mark up more schema), but they fail precisely on the markup that pays off most for a local business (we will see this in point 3). They pile up Organization, WebSite or BlogPosting from the template, and forget the LocalBusiness that actually moves the needle locally. The program met its goal of getting SMEs online, and its business impact is documented, with a +43 % rise in customers and +36 % in revenue among companies in the first call (BBVA / Red.es). But "being online" and "being technically optimized for local SEO" are two different things.

3. Do websites with structured data rank better? Only if it's the right kind

This is the study's central finding, and it has two layers.

Layer 1: having "some" schema gives you no advantage. The 230 sites with structured data rank practically the same (average position 2.79) as the 61 with none (2.85). At first glance, it seems schema doesn't help. And technically, that is what Google has been saying for years: structured data is not a direct ranking factor. What it does is enable rich results and help search engines understand the content (Google Search Central; Search Engine Journal).

Layer 2: the *type* of schema changes everything. When we isolate the 59 sites that specifically implement `LocalBusiness`, the picture clearly flips:

• Better average position: 2.47 versus 2.89 for the rest (median 2.0 versus 3.0).
• More SERP appearances: 1.46 versus 1.36.
• And, as a bonus, less security debt: 3.88 missing headers versus 4.37. They are simply better-tended sites across the board.
• Among Top 3 sites, the presence of LocalBusiness doubles that of the rest: 24.9 % versus 11.8 % for those ranked 4th and below.

That this advantage appears despite the selection bias we mentioned is exactly what makes it credible. It's not that the markup magically "pushes" the ranking. It's that LocalBusiness tells Google precisely what you are and where you operate, feeds the knowledge panel and rich results, and reinforces local relevance signals. The sector's reference literature places it among the most influential on-page elements for local business, alongside geo-targeted titles and the mobile experience (Whitespark, Local Search Ranking Factors; Moz, Local Search Ranking Factors).

The lesson is not whether you mark up, but what you mark up. A well-built LocalBusiness, with consistent NAP, hours, geo-coordinates and sameAs, is worth more than ten generic schema types inherited from the template.

4. Do more schema types equal a better position? No. Quantity is not the lever

If markup quality matters, does quantity? We measured it directly. The correlation between the number of schema types and position is essentially zero (−0.06), the same as with SERP appearances (−0.01). Grouping sites by markup volume (0, 1-2, 3-4, 5 or more types), the average position barely moves: 2.85, then 2.68, then 3.00, and finally 2.79. No trace of a clear trend.

Put another way: a site with 12 schema types does not rank better than one with 3 well-chosen ones. Piling up markup (Organization, WebSite, ImageObject, WPHeader, BlogPosting) without contributing a user-relevant entity is noise, not signal. This fits Google's official stance: schema enables search features, it does not guarantee their appearance, and unused markup causes no problems but also has no visible effect (Google Search Central).

The practical recommendation writes itself: fewer types, better chosen. For a training center, the core is LocalBusiness (or its HealthClub or ExerciseGym subtype), BreadcrumbList and, where relevant, FAQPage or Service. Everything else goes if it doesn't communicate a real entity.

5. Do security headers affect rankings? Not directly, but the data is alarming

The state of HTTP security headers in the sector is, plainly, bad:

• 190 of 291 sites (65 %) have all 5 evaluated security headers missing: HSTS, CSP, X-Frame-Options, X-Content-Type-Options and Referrer-Policy.
• Only 6 sites in total have the complete set.
• The average is 4.2 of 5 headers missing.

Impact on position? Practically nil in our sample (correlation −0.09 with position; zero with SERP appearances). In fact, sites missing more headers rank slightly better, probably because it's the large, established WordPress sites that neglect this area. And this is consistent with the documented reality: security headers are not a direct ranking factor, something Google has confirmed explicitly (Search Engine Journal). What is a ranking signal, since 2014, is HTTPS (Google Security Blog). And here the sector passes comfortably: 277 of 291 sites use HTTPS (95 %).

So why give it a whole section? Because technical SEO doesn't end at position. Security headers protect your users against clickjacking, MIME-type sniffing and referrer leakage. They reinforce trust. And above all, they are a reliable indicator of how much technical care sits behind a website. It's no coincidence that sites with flawless LocalBusiness have, on average, fewer missing headers. Security won't raise your rank, but its absence reveals the kind of neglect that does eventually take its toll.

6. Do robots.txt and sitemap.xml move the needle? They're the plumbing, not the engine

In this area we finally have a control group. 265 of 291 sites have `robots.txt` (91 %), 257 have `sitemap.xml` (88 %) and 234 declare the sitemap inside robots. That leaves 26 sites without robots and 34 without a sitemap to compare against. And the result? Sites without these files do not rank worse (2.62 on average versus 2.82). But there's a catch: those same sites are, almost all of them, minimal sites with nothing else (only 7.7 % of those without robots have structured data). It's not that skipping the sitemap helps; it's that they coincide with simple sites that rank for other reasons.

It's worth recalling what these files are and what they are not, per Google's own documentation.

robots.txt manages crawling, not indexing. It serves to avoid overloading the server and to guide the crawler. It is not a mechanism for hiding pages; that's what noindex is for. In fact, a blocked URL can still appear in results if it receives external links (Google Search Central).

The sitemap is a discovery and prioritization signal, not an "index now" button. It helps Google find new or updated pages, but it does not force crawling or improve ranking by itself.

In summary: neither robots.txt nor sitemap.xml are direct ranking factors (Google crawling documentation). They are infrastructure. But just as good plumbing doesn't make a house luxurious while bad plumbing ruins it, these files won't make you win, though doing them wrong can sink you. A misplaced Disallow can wipe you off the map. That most of the sector has them is good news. The next level is having them right.

The story the six headlines don't tell: the sector's technical hygiene is fragile

If you only look at the six questions, you might conclude that "the technical part doesn't matter because almost nothing correlates with position." That would be misreading the study. What the data actually screams is that the whole sector shares a weak technical foundation, and that is why no variable stands out. When everyone limps from the same leg, the technical part doesn't separate. Yet.

Look at the technical-hygiene X-ray. These figures come from the deep crawl of 106 sites out of the total, the subset we could crawl page by page:

• Images without an `alt` attribute: affects 79 of those 106 sites (75 %), averaging 22 images per site (max: 124). It's both an accessibility failure and a wasted image-SEO opportunity.
• Unsafe *cross-origin* links (target="_blank" without rel="noopener"): 73 sites, averaging 20, up to 223 on a single site. A security and performance risk, spread everywhere.
• Pages without a *meta description*: 57 sites. Pages without `H1`: 59 sites.
• Broken internal links (4xx): 42 sites, with one case of 141 broken links.
• Semantic `
  ` tag: present on 5 of the 291 sites (1.7 %). In a sector whose asset is being in a physical place, this is almost a joke.

And explicit local signals don't shine either, this time across all 291 sites: a tel: link on 50 % of sites, an embedded map on 51 %, WhatsApp on 29 %. The only signal with good coverage is the Instagram link (78 %), logical in such a visual sector. But social media is not local SEO.

This is, at heart, great news for anyone who wants to compete. Local SEO for training centers is not won today with heavy artillery. It's won with fundamentals almost nobody is tending. Marking up your address, your LocalBusiness, your images and your links properly puts you, today, ahead of three out of four competitors.

Why this will matter more and more

The selection bias in our sample hides a time bomb for the sector. Today, in a low-competition niche, "being online with WordPress and a sitemap" is enough to show up in the SERP. But the bar is rising.

Rich search and AI-based answer engines consume structured data to build their responses. Without clean LocalBusiness, your business is invisible to that new discovery layer. Page experience (Core Web Vitals) is already part of Google's ranking systems, as a modest but real factor, and as a tiebreaker when content is similar (Google Search Central). And as competitors who do tend to the technical part enter the field, today's common floor will become tomorrow's basement.

The technical foundation is an investment that compounds. It won't give you a spectacular leap on day one, but it makes everything else (content, links, reviews) perform better and withstand each algorithm update.

Recommendations for the sector

If you run, build or optimize websites for training centers, or for any local business, here is the priority order the data points to.

1. Implement `LocalBusiness` well, not lots of schema badly. It's the only on-page lever that shows a real advantage in our sample (2.47 versus 2.89 in average position). Include consistent NAP, hours, geo-coordinates, sameAs and the correct subtype (ExerciseGym, HealthClub). Delete the useless generic schema.
2. Treat HTTPS as non-negotiable and headers as hygiene. You already have HTTPS, almost certainly; add at least HSTS, X-Content-Type-Options and Referrer-Policy. You won't move up a position, but you'll close the gap that betrays neglect.
3. Fix the invisible hygiene. alt on every image, rel="noopener" on external links, a single H1 per page, unique meta descriptions and zero broken internal links. It's plumbing work, but it's what separates a cared-for site from an abandoned one.
4. Recover the `
   ` tag and the local signals. A real tel: link, a semantic address, an embedded map and, where applicable, WhatsApp. You're a local business: tell the machines so in their own language.
5. Don't touch `robots.txt` and `sitemap.xml`, except to review them. Confirm the sitemap is declared in robots, that there are no accidental Disallow rules, and that the sitemap only lists canonical, indexable URLs.
6. Measure and compare. The whole sector shares your same flaws. Audit, prioritize by impact, and become the cared-for exception before your competition does.

Conclusion

Question by question, the data seems to say that technical SEO doesn't move rankings in this sector. But that reading is a statistical trap. It doesn't move rankings because almost everyone does it equally badly. The one variable that stands out, LocalBusiness schema, does so precisely because few tend to it. That is the shape of technical SEO in local SEO: it's not the engine that launches you, it's the ground that holds everything else up. And in a sector where the ground is cracked, building on firm foundations is not a marginal advantage. It is the advantage.

Training centers understand one principle better than anyone: visible results are built on a base nobody sees. Their SEO should train the same way.

References

1. W3Techs / Automattic — WordPress Market Share (global CMS share). https://wordpress.com/blog/2025/04/17/wordpress-market-share/ · https://w3techs.com/technologies/details/cm-wordpress
2. Government of Spain — Kit Digital Program (España Digital). http://espanadigital.gob.es/lineas-de-actuacion/programa-kit-digital · https://www.acelerapyme.gob.es/en/kit-digital
3. BBVA / Red.es — What Kit Digital is and how to apply (impact on customers and revenue). https://www.bbva.com/es/es/empresas/que-es-y-como-solicitar-el-kit-digital-para-pymes-y-autonomos-en-espana/
4. Google Search Central — General Structured Data Guidelines (structured data is not a direct ranking factor). https://developers.google.com/search/docs/appearance/structured-data/sd-policies
5. Search Engine Journal — Google Confirms Structured Data Won't Make A Site Rank Better. https://www.searchenginejournal.com/google-confirms-that-structured-data-wont-make-a-site-rank-better/544433/
6. Whitespark — Local Search Ranking Factors (importance of LocalBusiness and local signals). https://whitespark.ca/local-search-ranking-factors/
7. Moz — Local Search Ranking Factors Study. https://moz.com/local-search-ranking-factors-study
8. Google Security Blog — HTTPS as a ranking signal (2014). https://security.googleblog.com/2014/08/https-as-ranking-signal_6.html
9. Search Engine Journal — Google Answers If Security Headers Offer Ranking Influence. https://www.searchenginejournal.com/security-headers-and-ranking-influence/488781/
10. Google Search Central — Robots.txt Introduction (crawling, not indexing). https://developers.google.com/search/docs/crawling-indexing/robots/intro
11. Google Crawling — How Google Interprets the robots.txt Specification. https://developers.google.com/crawling/docs/robots-txt/robots-txt-spec
12. Google Search Central — Understanding Google Page Experience / Core Web Vitals. https://developers.google.com/search/docs/appearance/page-experience