Supplement store · Shopify

WU Nutrition

I run online sales for a client, a supplement brand. That includes the store, the advertising that brings visitors, the numbers that show what works, and the AI assistant that helps sell. I set it up so their team can run it without depending on me.

Write to me See the product

WU Nutrition home page — Real screenshot of the project

Work for a brand: type of project
Shopify · Liquid: main tools
5: real screenshots of the project

01 · The challenge

What was going on

Context

WU Nutrition sells supplements through its own store and also on Amazon and Mercado Libre. The underlying challenge is that whatever an ad promises has to still be true when the person lands on the page, picks a product, pays, and comes back months later.

The problem

There were many separate initiatives, each solving its own piece. What was missing was pulling them into something reusable the team could move without every change becoming a new project.

My part

I run their ecommerce end to end. I propose where it is going, and I also do the work: design, code the store, dig into why people are not buying, run the advertising, and build the automations.

02 · How I solved it

How I decided to solve it

01
See the ad, the page it lands on, the catalog, the product, and the cart as one path the customer walks, instead of five pieces each person handles separately.
02
Keep the store blocks configurable, so the team launches campaigns and changes content without asking a developer for anything.
03
Use AI where it genuinely helps, always with real prices and data behind it and a person deciding at the end.

03 · What I made

What got built

A set of blocks for assembling campaign pages, the homepage, the catalog, and product pages without starting from scratch.
A side cart that shows how much is left to the next perk, suggests products, and always reflects real stock.
An assistant that answers buying questions, the automated customer emails, and the dashboard where you see whether any of it is working.

04 · Charts

The numbers as charts

WU-SYS

DTC experience system

Layers that should read as one journey, not isolated initiatives.

Acquisition and message20%
Paid media, creatives, and continuity into the destination.
Discovery (home / PLP)20%
Quiz, commerce navigation, and card-level signals.
Decision (PDP)25%
Gallery, trust, and sticky add-to-cart.
Conversational assistance20%
Daniela with server-verified commerce data.
Cart and CRM15%
Drawer, value progression, and Klaviyo.

Relative experience-system weights; not spend or commercial outcomes.

WU-AI

Assistant architecture

AI structures language; the server validates prices and promotions.

Entry / quick replies25%
Guided intents in the client.
Worker / guardrails30%
Scope limits and policy.
Verified commerce data30%
Prices and promos outside the model.
Generative response15%
Writing and structuring only.

Methodological flow diagram; no usage metrics are published.

05 · The product

Explore the project

Mobile home with the summer campaign, quiz banner, and access to Daniela.

01 Campaign, quiz, and assistance coexist without stealing the primary CTA.

Product grid with discounts, verified reviews, 60/120-cap variants, and 24–48h delivery.

02 Cards that sell: struck-through price, verified rating, variants, and visible logistics.

Daniela recommends PAZ and builds the 2x1 Calma Total combo with direct add to cart.

03 AI turns context into an offer: a 2x1 combo with visible savings and add-to-cart in chat.

UGC story with verified purchase and a product card to add without leaving the video.

04 Shoppable social proof: a real testimonial with an add-to-cart overlay.

Cart with progressive rewards bar, cross-sell, and secure-payment signals.

05 The cart grows the ticket: free shipping earned and the next reward in sight.

06 · How it turned out

What changed in the end

Launching a campaign stopped being a development job: it is assembled from blocks that already exist.
The store, the advertising, the emails, and the numbers stopped living in separate tools.

What I would not compromise on