PICK’D
A marketplace concept that makes buying from local farms as easy as grocery delivery.
PICK’D is a marketplace concept that shifts everyday grocery spending toward local farms and farmers markets. Shoppers pre-order mid-week from multiple vendors in one cart and pick up fresh on market day. Vendors get pre-orders and better forecasting instead of relying on walk-up sales alone.
I designed and built it solo for my Harvard course, Digital Media: From Ideas to Designs and Prototypes.
Who it’s for
Four kinds of users, one marketplace
- Shoppers who want local food on a flexible schedule
- Farms and food makers who want pre-orders before market day
- Market organizers who want digital tools instead of spreadsheets
- Experience hosts offering farm dinners, classes, and u-pick
Mapping the experience
Five journeys, every decision point
Browse to checkout
Filter products, open a detail overlay, adjust quantity, and check out in three steps: contact, delivery, payment.
Vendor discovery to trust
A shopper lands on a new vendor’s profile and decides whether to try them. Years at market and review count do the work of building trust.
Sold out recovery
When an item sells out, shoppers can accept a substitution, remove it, or message the vendor first. This took the most iteration, since it had the highest chance of ending in an empty cart.
First-time onboarding
Six steps covering account, location, interests, experiences, and dietary needs, with a step counter so people always know how much is left.
Shared vendor pages
A friend who isn’t signed up can still browse a shared vendor page fully. A gentle sticky “sign up to order” bar replaces a hard paywall.
Information architecture
Only as much navigation as each page needs
The site is organized around four main collections: Products, Vendors, Venues, and Experiences, plus overlays for cart and onboarding. A global nav holds search, favorites, and cart, with a secondary nav for the four content types.
Local navigation only shows up where it earns its place: filter chips and a sidebar on the product grid, and a step indicator in onboarding. Vendor and venue pages open in a single modal, so they don’t need anything extra.
Design system
Components, not one-off screens
I built a component library with defined states, variants, and accessibility notes for each piece. A few highlights:
- Substitution Selector. Lets shoppers pre-approve a backup item, with three states that stay readable at a glance inside a cart line. The hardest one to get right.
- Message Bubble. A simple buyer-to-vendor thread for questions like allergen checks before adding to cart.
- Vendor Group Header. Groups the cart by farm instead of a flat list, reinforcing that you’re buying from real people.
- Order Confirmation Card. Names every vendor in the order with a link back to their profile. Naming the farm is the real payoff of the product’s promise.
- Empty and error states. Each has its own illustration and copy for the specific dead end, instead of one generic “nothing here.”
Interaction and tone
Unhurried, tactile, and trust first
Motion is meant to feel calm, not app-store slick. Hover states shift color instead of size so nothing feels jumpy, filter chips swap color in 0.15 seconds, and menus open instantly. The hero’s rotating text is the one place motion is meant to be noticed.
Copy leads with where food comes from, and system messages like low stock or substitutions are written to ease the worry of ordering from a farm instead of a warehouse.
Accessibility
Checked while building, not after
Primary buttons pass WCAG AAA, and stock and season badges pass AA, always using a text label alongside color. Icon buttons and the quantity stepper land just under the 44pt tap target, which I flagged as a fix instead of shipping quietly.
Where it stands
A working prototype I’m proud of
PICK’D is a prototype, not a shipped product. Cart and checkout are wired to console logs instead of a live backend, and a couple of onboarding steps are still shells. It’s the project I’m proudest of for the structure and decision logic behind it, not just the visuals.
PICK’D
A marketplace concept that makes buying from local farms as easy as grocery delivery.
PICK’D is a marketplace concept that shifts everyday grocery spending toward local farms and farmers markets. Shoppers pre-order mid-week from multiple vendors in one cart and pick up fresh on market day. Vendors get pre-orders and better forecasting instead of relying on walk-up sales alone.
I designed and built it solo for my Harvard course, Digital Media: From Ideas to Designs and Prototypes.
Who it’s for
Four kinds of users, one marketplace
- Shoppers who want local food on a flexible schedule
- Farms and food makers who want pre-orders before market day
- Market organizers who want digital tools instead of spreadsheets
- Experience hosts offering farm dinners, classes, and u-pick
Mapping the experience
Five journeys, every decision point
Browse to checkout
Filter products, open a detail overlay, adjust quantity, and check out in three steps: contact, delivery, payment.
Vendor discovery to trust
A shopper lands on a new vendor’s profile and decides whether to try them. Years at market and review count do the work of building trust.
Sold out recovery
When an item sells out, shoppers can accept a substitution, remove it, or message the vendor first. This took the most iteration, since it had the highest chance of ending in an empty cart.
First-time onboarding
Six steps covering account, location, interests, experiences, and dietary needs, with a step counter so people always know how much is left.
Shared vendor pages
A friend who isn’t signed up can still browse a shared vendor page fully. A gentle sticky “sign up to order” bar replaces a hard paywall.
Information architecture
Only as much navigation as each page needs
The site is organized around four main collections: Products, Vendors, Venues, and Experiences, plus overlays for cart and onboarding. A global nav holds search, favorites, and cart, with a secondary nav for the four content types.
Local navigation only shows up where it earns its place: filter chips and a sidebar on the product grid, and a step indicator in onboarding. Vendor and venue pages open in a single modal, so they don’t need anything extra.
Design system
Components, not one-off screens
I built a component library with defined states, variants, and accessibility notes for each piece. A few highlights:
- Substitution Selector. Lets shoppers pre-approve a backup item, with three states that stay readable at a glance inside a cart line. The hardest one to get right.
- Message Bubble. A simple buyer-to-vendor thread for questions like allergen checks before adding to cart.
- Vendor Group Header. Groups the cart by farm instead of a flat list, reinforcing that you’re buying from real people.
- Order Confirmation Card. Names every vendor in the order with a link back to their profile. Naming the farm is the real payoff of the product’s promise.
- Empty and error states. Each has its own illustration and copy for the specific dead end, instead of one generic “nothing here.”
Interaction and tone
Unhurried, tactile, and trust first
Motion is meant to feel calm, not app-store slick. Hover states shift color instead of size so nothing feels jumpy, filter chips swap color in 0.15 seconds, and menus open instantly. The hero’s rotating text is the one place motion is meant to be noticed.
Copy leads with where food comes from, and system messages like low stock or substitutions are written to ease the worry of ordering from a farm instead of a warehouse.
Accessibility
Checked while building, not after
Primary buttons pass WCAG AAA, and stock and season badges pass AA, always using a text label alongside color. Icon buttons and the quantity stepper land just under the 44pt tap target, which I flagged as a fix instead of shipping quietly.
Where it stands
A working prototype I’m proud of
PICK’D is a prototype, not a shipped product. Cart and checkout are wired to console logs instead of a live backend, and a couple of onboarding steps are still shells. It’s the project I’m proudest of for the structure and decision logic behind it, not just the visuals.