The rule I broke
I fell in love with the solution, not the problem.
My rule is to fall in love with the problem, not the solution. With Bespoke I did the opposite. The real problem was artisans connecting with the people who’d buy their work. Both versions chased an idea that looked great, without checking that anyone wanted it.
Then and now
Start from the problem
Help artisans increase their sales and get discovered.
Today I’d start with the problem we skipped. Artisans with great work and their own style need more people to find them. So Bespoke would be storefronts for artisans and a discover feed of real pieces you scroll like Pinterest. You could search with a picture of something you like, say a mid-century chair, and find the artisans who make that kind of work. The AI would read their real work, not generate pieces that don’t exist.
Found by their work
- ready · press run, or hover any part of the system
What it would do
Four things, for artisans and the people who buy from them.
A storefront for each artisan. Their pieces, their story, and how they make their work, in one place they can share. They keep designing what they enjoy making, and they can open commissions when they want to.
A discover feed. Real pieces from many makers, browsed like Pinterest. Starting with one category, like ceramics, so the feed is worth scrolling from day one.
Search by image. Add a photo of something you like, and see the artisans whose real work looks like it. You buy a piece they’ve made, or start a conversation about one they’d want to make.
A certificate of origin. Every piece ships with one, made by the artisan in a few minutes. Buyers can check it with a code, with no wallet or app to learn.
Looking back
What I'd keep
The discover feed.
Browsing real work by the people who made it was the best part of the second version. Now it's the center.
What I'd change
A certificate, not a blockchain.
Each piece ships with a certificate of origin the artisan makes in minutes. Buyers have nothing new to learn.
What we never asked
Whose problem, and which category.
We never chose a problem or a category, and designed for how people might buy in ten years. Now I'd start with one.
How I'd build it
Build it, and let artisans tell me what's next.
In 2024 Bespoke stayed in Figma. Since then, working on vers1ons and in my internships, I’ve learned to use AI as leverage to build and to learn. Now I’d build a working version of this, put it in front of a few artisans in one category, and let what they and their buyers do decide what comes next.
What still holds
Handmade will matter more.
One-of-one handmade work will matter more as more things are made by machines, and buyers want to know where things come from. That read hasn’t changed.