Mary's Land Farm runs farm experiences across multiple locations: alpaca encounters, cooking classes, barn stays, weddings. The brief was to replace Roller, their booking backend.
Mapping how the team actually worked, I found seven tools doing the job of one, and no single view of the business. So I made the case to replace all seven with one platform.
Client
Mary's Land Farm
Role
Sole Designer
Scope
End-to-end admin platform
Year
2025–Present
Context & Research
Mary's Land Farm runs farm experiences across multiple locations: alpaca encounters, cooking classes, barn stays, weddings. The brief was to replace Roller, their booking backend.
Mapping how the team actually worked, I found seven tools doing the job of one, and no single view of the business. So I made the case to replace all seven with one platform.
Problem
A single guest message meant four context switches: read it in Wavelength, open Roller to find the booking, copy the details, switch back to reply.
The platform had to close those gaps, not compete with the tools.
Solution
Every tool-switch the team used to make became a panel instead.
Decision · 01
Every experience is an image card: photo, price, duration, capacity, tags. Publishing one auto-creates the customer-facing listing, replacing WordPress. Cards, not a data table: the team sees the product the way a guest does, which keeps decisions grounded.
Experience cards: price, capacity, tags
Decision · 02
235 bookings in one table: reference, customer, experience, check-in/out, guests, status, balance, source. The Source column (Web or Admin) replaced manual Excel tracking; filters, calendar view, and Add Booking are always one click away.
235 bookings, one table with Web/Admin source
Decision · 03
Guest SMS lived in Wavelength, so every reply meant four context switches. I built two-way SMS into the platform, with a guest-profile panel showing booking reference, experience, date, and status beside each conversation. Not in the original brief. Should have been.
Messaging with live booking context alongside
Decision · 04
The team ran automated emails through SendGrid but couldn't manage templates without technical help; broken formatting and wrong timing were routine. I built a simple editor: rich text, variable insertion, trigger rules ("24 hours before check-in"), and a live desktop/mobile preview. They went from filing a request to doing it themselves.
Template editor: variables, triggers, live preview
Decision · 05
Revenue, bookings, average order value, 12-month trend, today's schedule: one screen. Before, assembling that picture meant opening Excel, Roller, and Google Calendar separately. Quick Actions surface the four most-used tasks; no buried navigation.
Dashboard: the whole business, one screen
Decision · 06
The team needed answers about revenue, bookings, and payment status without rebuilding spreadsheets by hand every month. I built a report builder: pick a type, set a date range, choose the fields you need, preview, then export. One place to answer "how did we do?", no reconciliation, no spreadsheets.
Report builder: sales summary, fields, live preview
Impact
Seven tools replaced by one platform. Live across farm locations, 235+ bookings managed, every employee on one system.
7 → 1
Tools consolidated into one platform
235+
Bookings managed
Live
Used daily across farm locations
Reflection & Results
The biggest decision wasn't a UI call, it was the scope conversation.
They asked for a Roller replacement. I mapped their actual workflow first, found seven tools doing the work of one, and made the case to consolidate. That decision changed what we built.
The messaging panel is where it shows most clearly: a farm host reads a guest message and sees the full booking context (experience, date, status, reference) without leaving the screen. That's the standard every feature gets measured against.
