53% of Colombian users abandon a site that takes more than 3 seconds to load. Every additional second of loading time reduces conversions by 7% to 12%. If your site takes 6 seconds to load on mobile—which is the average for unoptimized WordPress sites in Colombia—you’re losing more than half of your visitors before they even see a single line of content.
A site that loads in 3 seconds versus 1 second converts up to 3 times less. It’s not a technical problem—it’s a sales problem.
Over 75% of Colombian web traffic comes from mobile devices. Mobile speed isn’t optional—it’s your user’s primary experience.
Google has used Core Web Vitals as a ranking factor since 2021. Failing LCP or CLS directly affects mobile search rankings.
Website speed is the technical issue with the most immediate impact on Colombian businesses—and at the same time, the most overlooked. I’ve seen companies invest 5 million COP per month in Google Ads to drive traffic to a site that takes 8 seconds to load on mobile. The advertising budget results in users arriving, waiting, getting frustrated, and leaving. All that money spent on ads funds user frustration, not conversions.
This guide explains what Google measures, what metrics you should aim for, what causes the most common speed issues in Colombia, and how to fix them. If you’d like us to handle this for you, at Webhexup we manage technical SEO optimization—including Core Web Vitals—for Colombian websites.
The three metrics Google uses
to evaluate your site’s speed.
Time until the largest element on the page (main image or block of text) is visible. This is the most important metric for the user's perception of speed.
The site’s response time when the user clicks or taps the screen. A site with heavy JavaScript blocking the main thread fails here even if it loads quickly visually.
Visual stability—how much page elements shift as the page loads. If text jumps out of place when an image appears, Google records it as a high CLS.
Google evaluates these metrics under real-world user conditions using data from Chrome (CrUX — Chrome User Experience Report). It’s not just the PageSpeed Insights result on your computer—it’s the actual load average for all users who visited your site in the last 28 days from their actual devices in Colombia.
How to diagnose your site’s speed
in 5 minutes.
pagespeed.web.dev — Google’s official tool. It displays a score from 0 to 100, real-world vs. lab Core Web Vitals metrics, and opportunities for improvement prioritized by impact. Always check the “Mobile” tab—that’s the one Google uses for ranking.
"Page Experience" report → "Core Web Vitals." Shows which specific URLs on your site have actual issues measured among real users—not just lab estimates. This is the data source Google uses for ranking.
gtmetrix.com — lets you view the full loading cascade, identify which specific resource is blocking rendering, and measure performance from different locations. Set the measurement location to Latin America for results that are more representative of Colombian traffic.
webpagetest.org — the most comprehensive tool for advanced technical diagnostics. It allows you to measure with 3G/4G connection throttling (representative of the average user in Colombia) and generates a visual filmstrip showing how the page loads millisecond by millisecond.
The most common causes of slow
performance on Colombian websites.
| Cause | Impact on LCP | Frequency in Colombia | Difficulty in Fixing |
|---|---|---|---|
| Images that are neither compressed nor converted to WebP | Very high — may account for 60% of the file size | Very high | Low |
| Slow hosting without server caching | Very high — high TTFB blocks everything | High | Medium (requires migration) |
| Blocking third-party JavaScript | High — delays DOM rendering | High | Medium |
| Unconfigured cache plugin | High — cache inactive or misconfigured | Very high | Low |
| No CDN — single server in the U.S. or Europe | Medium-high — high latency from Colombia | High | Low (free Cloudflare) |
| Google Fonts without `display=swap` | Medium — blocks text rendering | High | Very low |
| Unused CSS loading on every page | Medium — unnecessary weight on every load | Medium | Medium |
| Images without declared dimensions (CLS) | High CLS — visual shifts during loading | High | Low |
How to Improve Website
Speed Step by Step — by Impact.
Convert all images to WebP and add dimensions
WebP reduces image file sizes by 25% to 60% compared to JPG/PNG while maintaining the same visual quality. In WordPress: ShortPixel or Imagify handle the conversion automatically. Additionally, adding the attributes width and height to each image in HTML eliminates CLS caused by images without declared dimensions. Impact: LCP -40% on sites with large images.
Configure LiteSpeed Cache or WP Rocket correctly
Simply installing the plugin isn’t enough—the configuration makes all the difference. Enable: full-page caching, CSS and JS minification, image lazy loading, font preloading, and deferring non-critical JavaScript. With LiteSpeed Cache properly configured, most WordPress sites in Colombia go from a score of 40–50 to 85–95 on mobile without changing hosting.
Enable Cloudflare as a CDN
The free version of Cloudflare reduces latency for Colombian users by serving site assets from nearby servers (it has nodes in Bogotá and São Paulo). It also adds DDoS protection, automatic Brotli compression, and edge caching. Impact: TTFB reduced by 40% to 60% for users in Colombian cities far from the main server.
Free — maximum ROI per hour investedRemove or defer third-party JavaScript
Live chat scripts, marketing pixels, social media widgets, and misconfigured analytics scripts block the browser’s main thread and directly impact INP/TBT. Audit using Chrome DevTools → Coverage tab to identify which JS is not used during the initial load. Defer non-critical scripts using the defer attribute or move them to the footer.
Add preload to the LCP image
The image that accounts for the LCP (usually the hero banner or the main image of the article) must be loaded with the highest priority. Adding <link rel="preload" as="image" href="hero.webp"> to the <head> section tells the browser to load this image before any other. It’s the most straightforward and lowest-risk LCP fix. It can reduce the LCP by 0.5 to 1.5 seconds with a single line of code.
Hosting is the foundation of everything:
the speed ceiling you cannot exceed.
No amount of cache, CDN, or code optimization can compensate for slow hosting. TTFB (Time to First Byte—the time it takes for the server to start responding) is the speed factor that depends exclusively on hosting and blocks all other optimizations.
| Hosting | Average TTFB | Server | PHP | Recommended for |
|---|---|---|---|---|
| Hostinger Business | 180–280 ms | LiteSpeed | 8.2+ | Colombian SMEs — best price/performance |
| SiteGround GrowBig | 160–240 ms | Nginx + SG Optimizer | 8.2+ | Medium-sized sites with variable traffic |
| Kinsta | 120–200 ms | Google Cloud C2 | 8.2+ | High-traffic projects on a budget |
| Generic shared hosting in Colombia | 800 ms–2.5 s | Unoptimized Apache | 7.x | Not recommended for SEO |
If your TTFB in PageSpeed Insights is greater than 600 ms, hosting is the primary issue—no amount of plugin or image optimization will solve the root cause. Check the "Time to First Byte" metric in GTmetrix in isolation from the rest of the load. A TTFB of less than 300 ms is the target for Colombian websites on quality shared hosting.
"You won’t notice a one-second difference in load time when browsing from your office with fiber-optic internet. But a user in Bucaramanga on mobile data will notice it. That user is your customer."
— José Fernando Muñoz · Webhexup
From 8 seconds to 1.8 seconds:
without redesigning the site.
Fashion e-commerce — Medellín · WooCommerce + Elementor
Initial status: LCP 8.2s · TBT 2,840ms · CLS 0.42 · Mobile PageSpeed: 22. Colombian shared hosting. 180+ uncompressed PNG images. 3 active caching plugins in conflict.
- Step 1: Migration to Hostinger Business with LiteSpeed — TTFB reduced from 1.8s to 210ms
- Step 2: Converting all images to WebP with ShortPixel — total page weight reduced from 8.4MB to 1.9MB
- Step 3: Removed 2 conflicting cache plugins, configured full LiteSpeed Cache
- Step 4: Cloudflare enabled + hero image preload + local Google Fonts
- Step 5: Dimensions specified for all product images — CLS from 0.42 to 0.04
Result: LCP 8.2s → 1.8s · TBT 2,840ms → 0ms · CLS 0.42 → 0.04 · Mobile PageSpeed 22 → 91. Implementation time: 2 days. Impact on sales: +34% mobile conversion rate in the following 30 days.