Let’s talk

Back to writing

Development

How I’d investigate a slow Shopify storefront

First I find out whether customers are waiting for the page, waiting for a button, or watching the page move under them.

“Can you make the site faster?” is a fair question. I usually follow it with: which page, on what device, and what feels slow? A homepage banner that takes ages to appear is a different problem from a cart button that sits there after I tap it.

Core Web Vitals help put a name to each problem. LCP is about how soon the largest visible content appears. INP is about responsiveness after interactions. CLS measures unexpected layout movement. Google’s good targets at the 75th percentile are up to 2.5 seconds for LCP, 200 ms for INP, and 0.1 for CLS. Shopify’s reports use real sessions over a rolling 30-day period, which I’d use to choose where to investigate first.

One symptom, one trace

If the hero is late, I’ll check when the browser discovers the image, how large it is, and what has to happen before it can paint. If a variant selector feels slow, I’ll record the interaction and look for JavaScript work that delays the next frame. If the page jumps when an image or embed appears, I’ll see whether the layout reserved enough space for it.

A LinkedIn post by Poptin’s Tomer Aharon described its app’s p75 INP moving from 280 ms to 120 ms after performance work. That’s one vendor’s result. What I take from it is that app code deserves a look when the trace points there.

I’d change one meaningful thing, repeat the same test, and then watch the field data after launch. I’d also check the feature still works. A faster cart button should still add the right item and tell me it’s there.

Have something in mind?

Let’s talk.

Tell me what you’re building. Let’s find the right way to bring it to life.