Updated 14 October 2026
Case study: improving the platform’s mobile experience
Saved homepage measurements after media and font improvements, with lab performance separated from search rankings and sales.
The problem and changes
The previous introduction video was 8,871,111 bytes. We prepared a 334,090-byte mobile version, approximately 96.2% smaller as a file, with a separate desktop source. Fonts moved to local hosting. Video startup handlers were consolidated, playback controls were retained, and reduced-motion and data-saving preferences were considered.
This percentage measures file size, not page-load time or conversion improvement.
Available measurements
- Source: saved PageSpeed Insights reports from 13 October 2026, 22:54 Cairo time, for release 7.0.3.
- Arabic mobile homepage: performance 98/100, LCP around 2.4 seconds and CLS around 0.008.
- English mobile homepage: performance 99/100, LCP around 1.8 seconds and CLS around 0.038.
- Browser checks on 14 October at a 390-pixel viewport: the mobile video played muted and inline, with no horizontal overflow on either homepage.
Lab results vary across runs, devices and networks. The reports did not provide sufficient visitor field data. These scores do not establish a Core Web Vitals pass across every page.
How to test the approach
- Record the URL, date, device configuration and page release before changing anything.
- Identify the resource that delays the initial experience. Do not reduce essential image quality purely to save bytes.
- Change a defined factor and repeat the same measurement procedure.
- Check muted playback, controls, reduced motion, language direction and forms on a real phone.
- Review field data when available and follow request completion and quality. Speed and sales are different outcomes.
Limits and maintenance
This is an implementation study on our platform, not a customer testimonial or independent comparison of businesses. No sales increase or search leadership is attributed to it because neither was established. The 14 October check confirmed the published HTML matched the release associated with the saved reports; later measurements can still differ.
Prepared by the platform team with AI assistance. Measurements and limits are disclosed without an independent reviewer claim.