I build Shopify and custom storefronts for small brands: memberships and subscriptions, booking, customer dashboards, and the admin tooling behind them. Nothing is built on your live site. Shopify work happens on an unpublished copy of your theme, custom builds on a staging site only you can reach, and either way it goes live when you say so.
Two, and that is deliberate. I would rather go deep on a short list than claim to cover everything.
Shopify and custom storefronts. Memberships and subscriptions, booking and scheduling, customer dashboards, and the admin tooling that keeps them running. You end up with a store your customers can use and you can manage without calling anyone.
AI put where it removes work rather than where it makes a headline: support replies, customer onboarding, content production, internal tools. The measure is hours you stop spending on the parts of the business that repeat.
Each project is written the same way: what the situation was, what could have gone wrong, and how the work comes back out if you want it gone.
Shopify theme development
A bike shop's helmet collection rebuilt to match a premium reference design, with filtering and sorting untouched. It switches on for one collection and off again from the Shopify admin.
Self-directed build on my own Shopify development store. No client, no customers, and no performance numbers to report.
Most of the risk in store work is not the code. It is what happens to a business while the code is being changed.
Changes are built on an unpublished copy of your theme, or on a staging site if the build is custom. Nothing a customer can see moves until you have looked at it and said go.
Features are built so they can be switched off without a code change. If something is not working for your business, removing it is a decision, not another project.
Including the options I turned down and why. You get the reasoning, not only the result, so whoever touches the store next is not starting from nothing.
Testing runs the whole checklist again after every change, not only the part that changed. Regressions get caught before you find them.
A mistake in my own process gets flagged and the process changes. That habit is the reason the four points above are true rather than aspirational.
I'm Felipe. ConvertiveCo is my practice, and it is one person. You talk to whoever is doing the work, from the first conversation to the handoff.
The habits on this page come from building web applications inside a large enterprise technology organization, where nothing ships without review and every change has to be explainable after the fact. Those constraints are why store work here is reversible and written down: at that scale you learn quickly that undoing a change matters as much as making it.
Before development, my work was on the marketing side: offers, audiences, and why customers actually buy. I don't sell that as a service, but it is in the room every time I decide how a product page, a signup flow, or a checkout should behave.
The work I want is the kind that outlasts the engagement. A store the owner can run, reasoning you can read months later, and a clear way to undo anything I built.
Before this practice was build-led, my work was creative strategy and messaging for e-commerce brands. These are speculative samples, written for brands that do not exist. They are here because that thinking still shapes how I build, not because they are on the menu.
Paid Social Scripts
A UGC-style paid social script pack built for a speculative cortisol-support supplement brand, designed to show strategy-led scripting, hook development, and paid social thinking across funnel stages.
Ad Concept Pack
A Meta ad concept pack for a speculative stress and cortisol-support supplement brand, built to show emotionally specific angle development for direct-response paid social.
Creative Strategy
A paid social creative testing matrix built to organize audience state, funnel stage, hook style, reframe, and testing priorities for a speculative supplement brand.
Tell me what you are trying to do and what you are worried about. If it is not something I should be the one building, I will say so.