Skip to main content
FHAVI

Free resource

Speed, measured the way a customer experiences it

Speed scores are easy to improve without making the store faster. This checklist is ordered by what actually delays a customer on a mid-range phone over a mobile connection — and it says which popular fixes are not worth the effort.

Measure first

Three numbers, on a real device

Lab scores and field data answer different questions. A lab score tells you what a fast connection sees; field data tells you what your customers actually experience. If they disagree, believe the field data.

Measure three things: time to first meaningful content on a mid-range Android over throttled 4G, total transfer size of the product page, and the number of requests before the page is usable. Then make one change and measure again — this list is not a redesign.

Group 1

Images, which are usually most of the problem

01

Checks

  1. Product images are served at the size they display — a 3000px file in a 600px slot is the most common single mistake.
  2. Modern formats (WebP or AVIF) with a fallback, which typically halves image weight at the same visual quality.
  3. Below-the-fold images are lazy-loaded, and the first hero image is not.
  4. Width and height are set so the layout does not jump as images arrive.
  5. Image CDN or platform resizing is in use rather than original files uploaded straight from a camera.

Group 2

Scripts and third parties

02

What to check

  1. Count the third-party scripts on a product page. Anything above ten is a decision, not an accident.
  2. Analytics and pixels load after the main content, not before it.
  3. Chat widgets, popups and review apps are deferred until interaction or idle time.
  4. Fonts are limited to two weights and served from the same origin or a fast CDN, with a system fallback.
  5. No render-blocking script sits in the head before the first paint.
03

What is usually a waste of time

  1. Chasing a perfect lab score while the field data says customers wait four seconds.
  2. Minifying an already-minified bundle while eleven scripts load before content.
  3. Switching hosts before fixing images — the same images will be slow anywhere.
  4. Preloading everything, which competes with the resources that actually matter.

Group 3

The mobile experience itself

04

Checks

  1. The add-to-bag action is visible without scrolling on a 360px-wide screen.
  2. No horizontal scroll on any page; tap targets are at least 44px.
  3. The sticky header does not consume more than about a tenth of the viewport.
  4. Forms use the right keyboard type and do not zoom on focus (font-size at least 16px).
  5. Checkout completes in one session on a phone, including the payment redirect and the return.
05

How to test it honestly

  1. Use a mid-range Android, not a desktop browser in device mode.
  2. Throttle to 4G and clear the cache between runs.
  3. Test on a product page, not the homepage — the homepage is always the fastest page you have.
  4. Test again after any theme or app change, because speed regresses silently.

Priority

If you only do three things

1. Serve correctly sized, modern-format images. It is usually the largest single win and it needs no redesign.

2. Remove or defer third-party scripts you cannot justify. Every one is a request, a parse and a risk.

3. Fix the mobile add-to-bag path before optimising anything else, because that is where the money is.

More resources

Keep going

The rest of the library

Every checklist here is free and ungated. The research pages carry the reasoning behind them.

All resources · Research library

Slow store and no idea where the weight is?

Send the product URL. We will tell you which of these is costing you most, and whether it is a theme problem or a script problem.

Prefer email? Write to md@fhavi.com