Speed is a revenue metric
Slow stores lose shoppers before anything is seen. That loss does not appear in your conversion rate as a problem — it appears as traffic that arrived and left, which most owners attribute to the wrong cause.
Improving load time is one of the few changes that improves every channel simultaneously, because it applies to every visitor regardless of where they came from.
Where the seconds go
| Cause | Share of the problem |
|---|---|
| Oversized images | Usually the largest single cause |
| Third-party scripts and apps | Frequently second |
| Theme code and page builders | Meaningful on older stores |
| Server response | Small on hosted platforms, large on self-hosted |
| Fonts and icon libraries | Small but easily avoidable |
Test the way customers experience it
- On a phone, on mobile data, away from your office
- The homepage, a collection page and a product page separately
- With a cleared cache, as a first-time visitor
- At a busy time of day, not at 7am
Testing your homepage on office fibre and concluding the store is fast is the most common measurement mistake in eCommerce.
Fix in order of effect
- Resize and compress every image over about 300KB
- Remove or defer scripts that do not need to load immediately
- Audit apps and remove what is unused
- Clean out leftover theme code from removed apps
- Then consider anything more involved
Measure the before and after
Record your load times before you change anything, then again after each change. Without that, you will not know which fix mattered and you will not be able to defend the effort.
Field data from real visitors matters more than laboratory tests, and most platforms and analytics tools now provide it.