Week 4
Live and taking orders
A custom Stripe-powered storefront where there had been none. One coffee, one hardcoded price, and a working checkout. Customers could buy online without arranging each order directly with the roaster.
Case 01 / Ecommerce Platform
From no storefront to an owner-managed catalog.
Mendivil Coffee is a roaster in Walnut, California that had no way to sell online — orders arrived by word of mouth. I built them a custom Stripe-powered storefront from nothing: live in four weeks, then owner-managed two weeks later. They now create, price, archive, and restore their own coffees, and customers buy several bags in one order with a grind chosen per bag.
- Zero to launch
- 4 weeks
- To owner-managed
- 6 weeks
01 — Shipping first, then stepping out
Selling in four weeks. Self-serve in six.
Week 6
Handed to the owners
The owners could create, price, archive, and restore coffees from their own admin. Customers gained a cart that keeps their selections between visits and a grind choice for each coffee. The admin also shows owners when catalog changes have reached the storefront.
The first launch deliberately hardcoded a single coffee to one Stripe price. That was the right place to stop: it got them selling weeks earlier, and price changes were rare enough that routing them through me cost almost nothing. The second stage removed me from that loop before the catalog grew enough to make it a real bottleneck.
02 — Decisions that mattered
Two choices that kept the store useful after launch.
Owners can change a price without rewriting history
Stripe prices cannot be edited in place. A price change creates a new active price and retires the old one, while existing orders keep what customers were charged. The owner sees one price field; the bookkeeping stays accurate without becoming their job.
Seasonal coffee can leave and come back safely
Discontinuing a coffee archives it instead of deleting it. Past order details stay intact, the product disappears from the catalog and checkout, and a seasonal roast can return without being rebuilt from scratch.
Building something similar?
Tell me what customers cannot do today and who needs to run it after launch. That is enough to start a useful conversation.
Discuss your project