← back to case study · live prototype →

Morsel: flow and assumptions

Two ways in, one menu. Before you go: tap a restaurant card for its menu, or a dish for its page and then its menu. At the table: pick the place, read its menu with photos where they exist, look closer at a dish, compare two or three, decide. The study covered the second. built marks what the prototype does. to test records an assumption identified during design. The completed usability study and follow-up questions document what the first round examined and what remains open.

Home: Nearby built

Restaurant cards with a photo strip and a coverage tag, then a grid of photographed dishes near the fixed area, nearest first. "I'm at a restaurant" sits at the top, one tap away.

→ Picker · Menu · Dish

Restaurant picker built

"Which restaurant are you at?" Photo cards with a coverage tag, filtered as you type. No location request. Places without menu data are listed below as "Menu not added yet" and are not tappable.

→ Menu

Menu built

The restaurant's sections in its order. Each row includes a photo or "No photo added yet: dish available", name, description, price and a compare toggle. Full menus have section chips and a coverage count. Partial menus explain what is missing directly before the first dish row.

→ Dish · All photos

All photos · Grid view built

Only photographed dishes, as a grid, with the count of dishes it leaves out and a link back to the full menu.

→ Dish · Menu

Dish built

Photos one at a time with a source label on each (Restaurant photo, Diner photo) and a counter. Then the menu's description, price, what the menu lists (labeled chips, or a sentence when it lists none of nuts, gluten, dairy or shellfish) and a one-line note that appearance can vary between visits: a photo helps picture the food but does not confirm ingredients or today's presentation.

→ Add to compare · View the full menu (or the restaurant’s dishes, for a partial menu) · Back

Compare built

One to three dishes side by side, from one restaurant or several: photo or an explicit missing-photo label, name, restaurant, price, description, what the menu lists and Remove. The grid wraps to two columns on narrow phones or with larger text. Returning from a dish restores the shortlist's scroll position and focus. Nothing is ordered.

→ Dish · Menu

Back and memory

Questions recorded before the study

AssumptionWhere it livesWhat would change if it fails
People at a table want to see the dishes on the menu in front of them to testThe whole direction. It comes from my own frustration, not from research.The product premise.
Browsing nearby photos is how people find a place, and it should lead into the menu rather than stand aloneHomeWhether Home is a feed or just the picker.
Menu order with visible gaps beats a photos-first orderMenu; missing-photo rowsThe default view, and how the Photos view is framed.
"No photo yet" reads as missing information, not as "not available"Menu rows, dish pageThe placeholder wording.
A source label changes how much a photo is trusted, and diner photos are still wantedDish photo strip, thumbnailsWhether diner photos are shown by default.
Two or three dishes are enough to decide betweenCompare limitThe shortlist size, or whether compare is needed at all.
People do not read a photo as proof of ingredients or portionThe photo note on the dish pageHow prominent the note is.

The study addressed the menu, missing-photo, source-label, compare and photo-note rows within its conditions. The first row, whether people at a table want this, and the Home row, that browsing nearby photos is how people find a place, were not tested in either round.

Supporting notes from the design and prototype review. Restaurants, menus and photos are fictional sample data.