R$690,000 recovered from a checkout leaking 90% of ready buyers.
Sofá na Caixa had the traffic. On its largest journey, 4,770 shoppers started checkout. 455 completed an order.
That is a 9.54% checkout completion rate on a high-ticket furniture purchase where the buyer had already decided.
Shopify Analytics comparison vs. prior equivalent period
What changed at Sofá na Caixa?
The engagement fixed a 16.5-second mobile homepage, stabilised a page that was shifting under the shopper, opened up product discovery, and rebuilt the mobile checkout experience where the largest leak sat.
The work generated R$690,000 in recovered revenue across the engagement.
Sofá na Caixa sells furniture on Shopify in Brazil. High ticket. Considered purchase. A buyer who needs confidence about fabric, size, delivery and payment before committing.
On the main journey (Home → Product Page → Cart → Checkout), 4,770 shoppers started checkout and only 455 finished.
Mobile was the worst of it: 8.37% of mobile checkout starters purchased, against 15.28% on desktop. And because mobile drove roughly five times more checkout starts, that gap was where nearly all the lost money sat.
The problem was not only checkout. The homepage took 16.5 seconds to reach its largest visual element and scored 14 out of 100 on performance. Only 14.7% of homepage visitors ever reached a product page.
Five leaks, from first impression to final click
Leak 1: The mobile first impression
16.5-second LCP. 14/100 performance score. Heavy scripts and popups delaying the first meaningful view.
Leak 2: Layout instability
Missing image dimensions, duplicate libraries, chat tools, trackers and non-essential widgets shifting the page under the shopper as it loaded.
Leak 3: Product discovery was closed off
Only 14.7% of homepage visitors reached a product page. Shoppers who reached a collection first showed much stronger product-to-cart behaviour: the path existed, the homepage just was not pointing at it.
Leak 4: Product pages did not build confidence before the CTA
Instalment and PIX payment information, fabric and size clarity, delivery estimates and trust signals were missing at the point where a shopper decides whether a high-ticket sofa is safe to buy.
Leak 5: Mobile checkout
The largest leak. 8.37% completion against 15.28% on desktop, carrying five times the volume.
Fix the first second, then the last click
1. Fixed the mobile first impression
Hero loading, image optimisation, deferred scripts, and corrected popup timing to eliminate initial load friction.
2. Removed layout instability and script bloat
Added explicit image dimensions, removed duplicate libraries, and cut non-essential widgets and redundant trackers.
3. Opened up product discovery
Rebuilt the homepage to guide shoppers into clear categories and product paths instead of leaving them to search.
4. Built confidence on the product page
Instalment and PIX options surfaced upfront. Fabric, size and delivery expectations made explicit. Stronger imagery and visible trust signals placed before the CTA.
5. Rebuilt the mobile checkout
Validated pricing, payment, shipping and form friction end to end, then removed what was causing checkout starters to abandon.
The result
Direct impact on store revenue
Lifted from 9.54% baseline
Faster LCP and clean path
Same traffic. Same ad spend. A buying path that finally held.
How this was measured
The R$690,000 is incremental revenue recovered, measured in Shopify Analytics against the previous equivalent period.
This is a before-and-after comparison, not a controlled test. Traffic volume did not support one, and the changes shipped together, so no single fix owns the result.
No promotion, launch or seasonal event fell inside either period.
From the founder
"He fixed the whole buying path for our sofas. Cleared up the product pages. Made the offer clear. Tested the pricing with real numbers. We made R$690,000 more."
Rubens Stuque · Sofá na Caixa
The last click is not where checkout is lost
A checkout leaking 90% of its starters is rarely a checkout problem alone. It is a shopper who arrived unsure, waited 16 seconds, could not find what they wanted, never got an answer about delivery or payment, and finally ran out of patience at the step where the money changes hands.
Fix the confidence upstream and the last click stops being the hardest one.
More case studies
