Which speed metrics affect your customers and your ranking, and the fixes that produce the biggest improvement.
Speed advice online is mostly noise about scores. What matters commercially is narrower than the tools suggest, and the fixes are less exotic.
The metrics that count
| Metric | What it measures | Target |
|---|---|---|
| Largest Contentful Paint | When the main content appears | Under 2.5 seconds |
| Interaction to Next Paint | How fast the page responds to a tap | Under 200 milliseconds |
| Cumulative Layout Shift | How much the page jumps while loading | Under 0.1 |
Chasing a perfect score in a testing tool is not the goal. These three, measured on a real phone on mobile data, are.
The fixes that produce the most improvement
- Images. Almost always the biggest problem. Serve modern formats, size them correctly, and never ship a 4000 pixel photo scaled down in CSS.
- Fonts. Each custom font is a blocking download. Two weights is usually enough, and self hosting beats a third party font service.
- Third party scripts. Chat widgets, analytics, tracking pixels and embeds are frequently the slowest thing on a page. Audit what is actually earning its place.
- Hosting. Cheap shared hosting has slow server response times, and no amount of front end work fixes that.
The New Zealand specific part
If your host is offshore, every request crosses the Pacific. A CDN with a local presence, or hosting closer to your audience, cuts a meaningful amount of latency before you optimise a single image. Check where your server actually is, because plenty of New Zealand agencies host in the US without mentioning it.
How to measure honestly
Test on a mid range Android phone on mobile data, not on your laptop on office wifi. The gap between those two experiences is enormous, and your customers are mostly having the first one.