Pearparadise is a Shopify store that sells fine pearl jewelry.
This is a different case study, as after showing why the metrics were bad, the client decided to not pursue green metrics.
But I still achieved a ~7% improvement.
Pearparadise wanted to improve their mobile metrics, which were failing for both INP(Interaction to Next Paint) and LCP(Largest Contentful Paint).
But it was failing only in non-customers countries with low speed connection, as the graphic below shows for LCP:
The client didn’t know it. After showing the data, he decided to save and not pursue the overall(origin) green metrics threshold, as the effort wouldn’t make sense.
The ~7%(1.6 secs to 1.5 secs) improvement in Largest Contentful Paint desktop improvement came from optimizing the client ads:
Update October 2026
The client limited access to certain areas and managed to green metrics:
