From 1–2 sales/day to 4–6 with R$100
From 1–2 sales/day to 4–6 with R$100
Jewelry e-commerce that grew ROAS by 121% through strategic repositioning, checkout optimization, and CRM. From zero to break-even in 18 months.
3.1%
Conversion Rate
+2.3x
ROAS
+79%
Checkout Conversion
+29%
Average Order Value
21%
Repeat Purchase in 30 days
-40%
Repetitive Support Questions
3,1%
Conversion Rate
+2,3x
ROAS
+29%
Average Order Value
21%
Repeat Purchase in 30 days
Context
Context
What it was:
Essavie was a DTC e-commerce brand for 925 silver jewelry that I founded and operated from May/2023 to Nov/2024.
Budget: R$100/day on Meta Ads.
Team: me (product, design, CRM) + paid traffic agency (contractor) + graphic designer (contractor).
What it was:
Essavie was a DTC e-commerce brand for 925 silver jewelry that I founded and operated from May/2023 to Nov/2024.
Budget: R$100/day on Meta Ads.
Team: me (product, design, CRM) + paid traffic agency (contractor) + graphic designer (contractor).
The problem:
Low conversion despite qualified traffic.
Users viewed the product, hesitated, and abandoned.
Confusing checkout.
Post-purchase anxiety.
Wrong brand positioning.
The problem:
Low conversion despite qualified traffic.
Users viewed the product, hesitated, and abandoned.
Confusing checkout.
Post-purchase anxiety.
Wrong brand positioning.
What I did:
I repositioned the brand from “traditional luxury” to “accessible luxury.”
Redesigned the landing page focused on the hero product.
Optimized checkout with A/B tests. Built post-purchase CRM.
What I did:
I repositioned the brand from “traditional luxury” to “accessible luxury.”
Redesigned the landing page focused on the hero product.
Optimized checkout with A/B tests. Built post-purchase CRM.
Result
Scaled from 1–2 to 4–6 sales/day, reached break-even, and decided to pause when unit economics showed I would need capital (R$100k) to scale.
Result
Scaled from 1–2 to 4–6 sales/day, reached break-even, and decided to pause when unit economics showed I would need capital (R$100k) to scale.
The Problem - User vs Business
The Problem - User vs Business
Lack of trust. Fear of scams.
Perception that “cheap = fake.”
Lack of trust. Fear of scams.
Perception that “cheap = fake.”
“Is this real silver or costume jewelry?”
“Do you have a physical store? Has anyone actually received it?”
“Why is it so cheap?”
Increase conversion and AOV without increasing ad budget.
Increase conversion and AOV without increasing ad budget.
0.7% conversion on landing page (below jewelry average)
High checkout abandonment (1.9% conversion)
Low average order value (R$198)
Very little repeat purchase (8% in 30 days)
100% dependence on paid traffic
Solution (What I did)
Solution (What I did)
1. Brand Repositioning
1. Brand Repositioning
Pivoted from “traditional luxury” (competing with Vivara/Tiffany) to “aspirational accessible luxury”.
What changed:
New visual identity (minimalist, sophisticated)
New naming: Estavie Jewelry → Essavie
Copy focused on self-esteem and enjoy the present moment
Photography that made silver look like white gold
Result:
+200% landing page conversion (0.7% → 2.1%)
Pivoted from “traditional luxury” (competing with Vivara/Tiffany) to “aspirational accessible luxury”.
What changed:
New visual identity (minimalist, sophisticated)
New naming: Estavie Jewelry → Essavie
Copy focused on self-esteem and enjoy the present moment
Photography that made silver look like white gold
Result:
+200% landing page conversion (0.7% → 2.1%)
2. Single-Product Landing Page
2. Single-Product Landing Page
Created a page dedicated to the hero product (best seller, highest margin, lowest return rate).
Decisions:
Hero image on white background + lifestyle shots
Price and installments above the fold
Direct CTA: “Buy” (not “Add to Cart”)
REMOVED: customer reviews (A/B test showed they created doubt)
CTA A/B Test:“Buy” (opens cart) → 3.4% conversion, R$256 AOV
“Add to Cart” (notification) → 2.1% conversion, R$268 AOV
Chosen: “Buy” (impulse > hesitation)
Result:
3.1% final conversion
Created a page dedicated to the hero product (best seller, highest margin, lowest return rate).
Decisions:
Hero image on white background + lifestyle shots
Price and installments above the fold
Direct CTA: “Buy” (not “Add to Cart”)
REMOVED: customer reviews (A/B test showed they created doubt)
CTA A/B Test:
“Buy” (opens cart) → 3.4% conversion, R$256 AOV
“Add to Cart” (notification) → 2.1% conversion, R$268 AOV
Chosen: “Buy” (impulse > hesitation)
Result:
3.1% final conversion
3. Simplified Checkout + Strategic Upsells
3. Simplified Checkout + Strategic Upsells
I couldn’t change the number of steps (Nuvemshop limitation), but optimized each step.
What I did:
Clear microcopy on shipping/payment/warranty
CTA always visible
Upsell only at payment step (didn’t compete with main decision)
Subtle cross-sell in cart
Tests:
Large free shipping banner → worsened conversion (distraction)
Gift in cart → worsened conversion (created distrust)
Small upsell at checkout → +3% AOV
Result:
Checkout 3.4% conversion (+79%)
AOV R$256 (+29%)
I couldn’t change the number of steps (Nuvemshop limitation), but optimized each step.
What I did:
Clear microcopy on shipping/payment/warranty
CTA always visible
Upsell only at payment step (didn’t compete with main decision)
Subtle cross-sell in cart
Tests:
Large free shipping banner → worsened conversion (distraction)
Gift in cart → worsened conversion (created distrust)
Small upsell at checkout → +3% AOV
Result:
Checkout 3.4% conversion (+79%)
AOV R$256 (+29%)
4. Post-Purchase and CRM
4. Post-Purchase and CRM
Automated communication to reduce anxiety and generate repeat purchases.
What I implemented:
Email flow: confirmation → shipping → delivery → review → repurchase (7/15/30/45 days)
WhatsApp Business: proactive support
Live chat (Chatra): real-time support
Result:
38% email open rate
21% repeat purchase in 30 days
-40% repetitive questions
Automated communication to reduce anxiety and generate repeat purchases.
What I implemented:
Email flow: confirmation → shipping → delivery → review → repurchase (7/15/30/45 days)
WhatsApp Business: proactive support
Live chat (Chatra): real-time support
Result:
38% email open rate
21% repeat purchase in 30 days
-40% repetitive questions
Process (What I did, how I did it)
Process (What I did, how I did it)
Research
10+ conversations on WhatsApp/Instagram analyzing recurring doubts
Hotjar: session recordings, heatmaps, click points
GA4: funnel analysis, drop-off, time on page
Benchmark: Vivara, Rommanel, indie brands (Rachel De La Mer)
Research
10+ conversations on WhatsApp/Instagram analyzing recurring doubts
Hotjar: session recordings, heatmaps, click points
GA4: funnel analysis, drop-off, time on page
Benchmark: Vivara, Rommanel, indie brands (Rachel De La Mer)
Wireframing & Ideation
Platform limitations forced me to test three predefined landing page structures: hero product, catalog, and brand storytelling. The hero product won.
Wireframing & Ideation
Platform limitations forced me to test three predefined landing page structures: hero product, catalog, and brand storytelling. The hero product won.
A/B Testing
15+ tests run:
CTA button (“Buy” vs “Add to Cart”)
Review placement (kept vs removed)
Free shipping banner (large vs small vs none)
Upsell (cart vs checkout vs post-purchase)
A/B Testing
15+ tests run:
CTA button (“Buy” vs “Add to Cart”)
Review placement (kept vs removed)
Free shipping banner (large vs small vs none)
Upsell (cart vs checkout vs post-purchase)
Methodology
1 week per test, minimum 100 sessions per variant.
Methodology
1 week per test, minimum 100 sessions per variant.
Insight
Insight
“The turning point was realizing that users valued security more than visual sophistication.”
“The turning point was realizing that users valued security more than visual sophistication.”
My Role
My Role
Everything, end to end.
Founder and operator
Product designer (branding, UX, UI, CRM)
Analyst (GA4, Hotjar, Stripe, Meta Ads)
Customer support (WhatsApp, email, chat)
Everything, end to end.
Founder and operator
Product designer (branding, UX, UI, CRM)
Analyst (GA4, Hotjar, Stripe, Meta Ads)
Customer support (WhatsApp, email, chat)
Collaboration:
Paid traffic agency: weekly meetings, creative alignment
Graphic designer: rebranding (I briefed, he executed, we decided together)
Influencer: freelancer, weekly meetings about Instagram content and creatives
Nuvemshop devs: occasional technical support
Collaboration:
Paid traffic agency: weekly meetings, creative alignment
Graphic designer: rebranding (I briefed, he executed, we decided together)
Influencer: freelancer, weekly meetings about Instagram content and creatives
Nuvemshop devs: occasional technical support
Learnings
Learnings
1. Clarity > Sophistication
For mid-ticket purchases, trust comes from clear communication, not complex UI.
2. Small optimizations compound
Each improvement (button, copy, upsell) was modest. Together, they generated 121% ROAS.
3. Test your assumptions
I was sure reviews would help. Testing showed the opposite.
4. Post-purchase is part of the product
CRM is not “extra,” it’s a core part of the experience.
5. Constraints force focus
A tight budget forced ruthless prioritization.
1. Clarity > Sophistication
For mid-ticket purchases, trust comes from clear communication, not complex UI.
2. Small optimizations compound
Each improvement (button, copy, upsell) was modest. Together, they generated 121% ROAS.
3. Test your assumptions
I was sure reviews would help. Testing showed the opposite.
4. Post-purchase is part of the product
CRM is not “extra,” it’s a core part of the experience.
5. Constraints force focus
A tight budget forced ruthless prioritization.
What I’d Do Differently
What I’d Do Differently
Test simplified landing pages BEFORE investing in full branding
Do more qualitative research (1:1 interviews) before assuming problems
Build a design system from day one to iterate faster
Test pricing strategies earlier
Why I Paused
Why I Paused
After Black Friday, I analyzed unit economics:
To scale, I needed ~R$100k in capital OR
Verticalization (own production) to improve margins
My goal was to validate my product design skills, not to build a large jewelry e-commerce while taking major risks.
Decision: apply these learnings to more complex and scalable products (fintech, AI, complex products).
After Black Friday, I analyzed unit economics:
To scale, I needed ~R$100k in capital OR
Verticalization (own production) to improve margins
My goal was to validate my product design skills, not to build a large jewelry e-commerce while taking major risks.
Decision: apply these learnings to more complex and scalable products (fintech, AI, complex products).
Tools
Tools
Design: Figma, Photoshop, paper and pen
Analytics: GA4, Hotjar, Meta Pixel
Platform: Nuvemshop (limited, but functional)
CRM: ActiveCampaign, WhatsApp Business, Chatra
Payments: Nuvem Pago (Pix, card, installments)
Ads: Meta Ads (Facebook/Instagram)