Average

Score Report

bellatavola.sk

Scanned on October 8, 2026

Unknown
jQuery 1.11.3
Font Awesome
openresty
Last scanned: October 8, 2026
AI verdict

Your site works well but has security gaps. Immediately enable HSTS on your server to protect customers. On the bright side, your SEO optimization is excellent.

Action plan

Action plan by impact and effort

Every problem found, sorted by how much fixing it helps and how much work the fix takes. Start with the quick wins.

How we rank →

Quick wins

5

High impact, little work. Do these first.

Bigger projects

9

High impact, but the fix needs planning.

+4 more

Small fixes

17

Lower impact, quick to do.

+12 more

Later

47

Lower impact and more work. Plan them once the rest is done.

+42 more

Performance

Weight: 25% · 8 passed · 8 warnings · 2 failed · 2 SPA (verify manually)

What we found: 482 KB wasted on unused code!

Why it matters: Over 200 KB of unused code significantly slows parsing and execution. This is one of the easiest performance wins — removing dead code requires no trade-offs.

Note: This check uses heuristic detection and may produce false positives. Manual verification recommended.

How to fix: You're loading 482 KB of code that isn't used on this page. 1) Audit plugins and remove unused ones, 2) Use code-splitting for page-specific JS, 3) Run PurgeCSS on your stylesheets.
Evidence:PSI-lab· Confidence: low

What we found: Only 19% of 16 scripts optimized — most are render-blocking

Why it matters: 13 render-blocking scripts can add 1-3 seconds to page load. Each synchronous script creates a sequential download-parse-execute chain.

How to fix: Add defer or async to all <script src='...'> tags. Render-blocking scripts are a leading cause of slow FCP. Defer maintains execution order, async does not.
Evidence:HTML-heuristic· Confidence: high

What we found: 220ms (good: <200ms)

Why it matters: TBT measures how long the main thread is blocked. During this time, clicks and taps don't respond — your store feels frozen. This directly impacts perceived quality.

How to fix: Reduce JavaScript execution: defer non-critical scripts, code-split large bundles, remove unused plugins. Third-party scripts (analytics, chat, ads) are often the biggest offenders.
Evidence:PSI-lab· Confidence: high

What we found: 2.5 MB, 101 requests (optimal: <1.5 MB, <50 req)

Why it matters: A 3 MB page takes 6+ seconds on 3G mobile. Most e-commerce traffic is mobile — heavy pages lose customers on poor connections.

How to fix: Reduce page weight: compress images to WebP (saves 30-50%), enable text compression (gzip/brotli), lazy load below-fold images, remove unused CSS/JS.
Evidence:PSI-lab· Confidence: high

What we found: 959 KB — over the 600 KB budget, but below the 1270 KB median of the award-winning shops we benchmarked

Why it matters: Measured on a first load with no interaction, so scripts gated behind cookie consent don't count against you here. The 600 KB bar is best practice, not the cohort average: the shops we benchmarked have a median of 1 270 KB, and the best of them (a category winner that won on mobile) ships 410 KB.

How to fix: Split the bundle so a product page doesn't load checkout and account code. Defer third-party scripts until interaction. Drop unused plugin bundles — on most shops these outweigh your own code.
Evidence:PSI-lab959 KB JavaScript· Confidence: high

What we found: 101 requests — over the budget of 80, but at or below the median of the award-winning shops we benchmarked (140)

Why it matters: The 80-request bar is best practice; the benchmarked cohort median is 140 and the best shop manages 37. Being under the median is not the same as being fast.

How to fix: Bundle CSS/JS that always loads together, serve icons as a sprite or font rather than individual files, and lazy-load below-fold images so they aren't requested on load.
Evidence:PSI-lab101 requests· Confidence: high

What we found: 166 images without lazy loading

Why it matters: Without lazy loading, ALL images download on page load — even those never scrolled to. On a product page with 20 images, lazy loading can save most of the initial download.

How to fix: Add loading='lazy' to all images below the fold. Keep the hero/LCP image eager-loaded. Native lazy loading is supported by all modern browsers.
Evidence:HTML-heuristic· Confidence: high

What we found: no-store, no-cache, must-revalidate, no-cache, private

Why it matters: no-cache/no-store forces browsers to re-download resources on every visit. Repeat visitors load your entire site from scratch every time.

How to fix: Set appropriate cache headers: static assets should have max-age=31536000 with versioned filenames. HTML pages can use max-age=0 with ETag for revalidation.
Evidence:HTTP-header· Confidence: high

What we found: 9 CSS files loaded

Why it matters: Each render-blocking CSS file adds network latency. Consolidating CSS from 8 to 2 files can save 200-400ms on first load.

How to fix: Consolidate CSS files by bundling them into 1-3 files. Each CSS file is a separate HTTP request that blocks rendering until downloaded.
Evidence:HTML-heuristic· Confidence: high

What we found: 9 CSS files without critical CSS extraction

Why it matters: Render-blocking CSS delays first paint. Inlining critical CSS eliminates the render-blocking round trip — the biggest FCP improvement for CSS-heavy sites.

How to fix: Extract critical above-the-fold CSS and inline it in <head>. Load remaining CSS asynchronously: <link rel='preload' href='styles.css' as='style' onload='this.rel="stylesheet"'>.
Evidence:PSI-lab· Confidence: high

10 issues found in Performance. Room for improvement.

Your store doesn't qualify for the Zulien Score Badge yet. Score 80+ to earn it!

📧 Re-scan reminders

We'll email you monthly when it's time to re-scan and track your improvements.

🚀 Improve your score

Our PrestaShop modules can fix most of these issues automatically.

Free 15-min consultation · No commitment

Share & Export

This is an automated heuristic estimate, not a professional audit. Findings reflect publicly observable signals at scan time and may include false positives.Learn about our methodologyReport inaccuracyRemove this report