Back

Scheduled Ordering System

Designed a restaurant scheduling system from zero — solving cross-role workflow failures and driving 53% adoption in the first month. Scaled to 16K+ restaurants and 2M+ customers.

Role

Senior Product Designer

Contribution

Led the strategy and design of a zero-to-one scheduling system across platforms. Drove cross-functional collaboration and mentored a junior designer.

Duration

1.5-month design · 4-month full delivery · 2021

Company

iCHEF (Restaurant Tech/B2B SaaS)

PROBLEM

The system didn’t support scheduled orders, leading to manual workarounds, lost sales, and cross-role operational confusion

Restaurants needed to accept future orders, but the system only handled right now. Owners worked around this by asking customers to write pickup times in order notes — creating a chain of errors across every role: wrong orders, mismatched staffing, and inaccurate revenue reports.

COMPLEXITY

Designing across 4 roles, 3 platforms, and diverse restaurant business models without disrupting existing operations

This wasn't a single feature. Every design decision had to work for four different people at once, support diverse restaurant operating models, and fit into workflows restaurants already relied on — without disrupting daily operations under legacy constraints.

BUSINESS IMPACT

Achieved 53% adoption in the first month, driving 1%+ online revenue growth

Launched to 2,000+ restaurants, achieving 53% adoption in the first month and driving 1%+ online revenue growth. The solution later scaled to 16K+ restaurants, serving 2M+ customers. UX improvements increased daily activation by 71%, while cross-functional alignment reduced development time by two weeks.

Adoption (First Month)

+53%

Online Revenue (First Month)

+1%

PROBLEM FRAMING

Research across both sides of the system — restaurants and customers — to define what a reliable scheduling system needs to do

Restaurant Research & Insights (B2B)

Restaurants have diverse workflows, but share the same core need: fewer missed orders, less effort

Through 100+ feedback entries and 8+ interviews, I found that despite diverse workflows, restaurants share the same core need — fewer missed orders and less manual coordination. These directly translate to revenue growth, operational flexibility, and efficiency — which became the core design priorities.

Shared Core Motivations Across Restaurants

💰

Prevent Missed Orders

Protect revenue from scheduling mistakes, off-hour loss

👥

Reduce Staff Workload

Less manual coordination and follow-ups

🍽️

Maintain Brand & Quality

Prepare orders at the right time

Customer Research & Insights (B2C)

Customers want a clear and low-effort way to schedule orders and pick up on time

Through benchmarking 4 ordering platforms, I found that customers need clear availability upfront and predictable pickup time selection — without confusion, extra steps, and drop-off. This directly shaped how I designed the ordering flow: making scheduling visible, familiar, and frictionless.

What Clear Scheduling Looks Like in Practice

Store Closed

Make scheduling availability visible upfront

Order Process

Make pickup time selection clear and predictable

DESIGN DECISION

Designing an end-to-end experience by making deliberate trade-offs across roles, platforms, and mental models

Merchant Platform for Restaurant Owners

Easy, flexible setup for diverse restaurant workflows

① One-click import Most restaurants schedule pickups within business hours. Rather than asking owners to re-enter existing data, I designed a one-click import that pulls in their current business hours directly — reducing setup friction and the risk of configuration errors.

② Navigation placement I initially placed the feature within the existing menu structure, but pilot testing showed owners couldn't find it. Moving it to a dedicated entry point drove 71% higher daily activation rate.

❌ Before
Rely on customers to manually specify pickup times in order notes
✅ After
Define pickup rules at the system level instead of relying on manual instructions
❌ Navigation placement Idea A: Reusing existing IA with low discoverability and high density
✅ Navigation placement Idea B: A dedicated entry point designed for clarity and future scale
❌ Before
Rely on customers to manually specify pickup times in order notes
✅ After
Define pickup rules at the system level instead of relying on manual instructions
❌ Navigation placement Idea A: Reusing existing IA with low discoverability and high density
✅ Navigation placement Idea B: A dedicated entry point designed for clarity and future scale
❌ Before
Rely on customers to manually specify pickup times in order notes
✅ After
Define pickup rules at the system level instead of relying on manual instructions
❌ Navigation placement Idea A: Reusing existing IA with low discoverability and high density
✅ Navigation placement Idea B: A dedicated entry point designed for clarity and future scale
❌ Before
Rely on customers to manually specify pickup times in order notes
✅ After
Define pickup rules at the system level instead of relying on manual instructions
❌ Navigation placement Idea A: Reusing existing IA with low discoverability and high density
✅ Navigation placement Idea B: A dedicated entry point designed for clarity and future scale

Online Ordering Website for Customers

A clear scheduling experience built on a single source of truth

Previously, restaurants asked customers to note pickup times manually, while the system generated a different time after checkout — causing errors and confusion. I unified pickup time into a single source of truth, enabling customers to choose between ASAP and scheduled pickup, with the selected time applied consistently across ordering, POS, and kitchen. I also replaced numeric dates with text-based dates to reduce formatting misunderstandings.

❌ Before
The system only supported ASAP orders, forcing customers write manual pickup notes, creating confusion across workflows.
✅ After
Pickup options are generated from restaurant-configuration, creating a single, reliable pickup time across workflows.
❌ Before
Conflicting pickup times from system output and manual notes caused confusion.
✅ After
A single, system-defined pickup time creates clarity, trust, and consistency.
❌ Before
The system only supported ASAP orders, forcing customers write manual pickup notes, creating confusion across workflows.
✅ After
Pickup options are generated from restaurant-configuration, creating a single, reliable pickup time across workflows.
❌ Before
Conflicting pickup times from system output and manual notes caused confusion.
✅ After
A single, system-defined pickup time creates clarity, trust, and consistency.
❌ Before
The system only supported ASAP orders, forcing customers write manual pickup notes, creating confusion across workflows.
✅ After
Pickup options are generated from restaurant-configuration, creating a single, reliable pickup time across workflows.
❌ Before
Conflicting pickup times from system output and manual notes caused confusion.
✅ After
A single, system-defined pickup time creates clarity, trust, and consistency.
❌ Before
The system only supported ASAP orders, forcing customers write manual pickup notes, creating confusion across workflows.
✅ After
Pickup options are generated from restaurant-configuration, creating a single, reliable pickup time across workflows.
❌ Before
Conflicting pickup times from system output and manual notes caused confusion.
✅ After
A single, system-defined pickup time creates clarity, trust, and consistency.

POS for Restaurant Staff

A minimal-disruption workflow for fast order management

① Workflow integration Restaurant staff rely on fast, familiar POS workflows. Early testing showed that large structural changes would slow down daily operations — so instead of redesigning the flow, I integrated scheduled orders into existing interaction patterns to minimize disruption.

② Print Later visibility Rather than hiding Print Later in settings, I made it visible and enabled by default for all restaurants. This maximized feature exposure across the board, while giving each restaurant the flexibility to configure it based on how their team actually works — supporting diverse restaurants without forcing a one-size-fits-all approach.

Flexibility Design
Configurable "Print-Later" action to support diverse restaurant workflows and maximize adoption
Order Confirmation View
❌ Before
The order confirmation view wasn’t aligned with staff workflows, slowing down order decisions.
✅ After
A workflow-aligned confirmation view enables faster, clearer order decisions.
Order Confirmation View Trade-Offs
Designed and tested multiple UI concepts to evaluate trade-offs and determine the final direction
Button Trade-Offs
Preserved familiar “Accept” behavior and added “Print Later” to support diverse workflows without retraining.
Flexibility Design
Configurable "Print-Later" action to support diverse restaurant workflows and maximize adoption
Order Confirmation View
❌ Before
The order confirmation view wasn’t aligned with staff workflows, slowing down order decisions.
✅ After
A workflow-aligned confirmation view enables faster, clearer order decisions.
Order Confirmation View Trade-Offs
Designed and tested multiple UI concepts to evaluate trade-offs and determine the final direction
Button Trade-Offs
Preserved familiar “Accept” behavior and added “Print Later” to support diverse workflows without retraining.
Flexibility Design
Configurable "Print-Later" action to support diverse restaurant workflows and maximize adoption
Order Confirmation View
❌ Before
The order confirmation view wasn’t aligned with staff workflows, slowing down order decisions.
✅ After
A workflow-aligned confirmation view enables faster, clearer order decisions.
Order Confirmation View Trade-Offs
Designed and tested multiple UI concepts to evaluate trade-offs and determine the final direction
Button Trade-Offs
Preserved familiar “Accept” behavior and added “Print Later” to support diverse workflows without retraining.
Flexibility Design
Configurable "Print-Later" action to support diverse restaurant workflows and maximize adoption
Order Confirmation View
❌ Before
The order confirmation view wasn’t aligned with staff workflows, slowing down order decisions.
✅ After
A workflow-aligned confirmation view enables faster, clearer order decisions.
Order Confirmation View Trade-Offs
Designed and tested multiple UI concepts to evaluate trade-offs and determine the final direction
Button Trade-Offs
Preserved familiar “Accept” behavior and added “Print Later” to support diverse workflows without retraining.

Order Prep Slip for Kitchen Staff

A single source of truth that eliminated conflicting pickup times

Previously, kitchen staff received order slips with two conflicting pickup times — one from the system, one from the customer's manual note. By establishing a unified scheduling system, the customer note became unnecessary, leaving only one system-defined pickup time on every slip. The kitchen problem solved itself.

Order Prep Slip for Kitchen

❌ Before

Conflicting pickup times from notes and system confused kitchen staff

✅ After

A single system-defined pickup time provides a clear source of truth for meal prep

SYSTEM DESIGN

Rebuilding the scheduling logic to eliminate conflicts across roles and platforms

Pickup Time Design

Defining pickup time rules to ensure clarity and a single source of truth

To prevent order confusion and lost sales, I rebuilt the pickup time logic across three key rules.

① Availability from pickup windows, not calendar dates — Prevents dates from appearing unavailable when restaurants are still accepting orders.

② ASAP is always the earliest valid option — Ensures customers never see a scheduled slot that appears sooner than ASAP.

③ ASAP time calculated at order placement — Prevents ASAP from appearing later than scheduled options. 93% of orders accepted within one minute made the trade-off low risk.

① Availability from pickup windows, not calendar dates

② ASAP is always the earliest valid option

③ ASAP time calculated at order placement

Legacy Design

Reducing system complexity and engineering overhead by sunsetting legacy auto-cancel logic

The 10-minute auto-cancel rule conflicted with long-term scheduling and increased system complexity. I led a cross-functional review to assess operational impact and risk. After validating real scenarios, we removed the rule without negative impact, simplifying the system and reducing engineering overhead.

Validation

Validating the new logic across real operating models

I tested the new rules across diverse restaurant models, including split shifts, overnight operations, and large scheduled orders. This ensured rule consistency, aligned cross-functional understanding, and prevented downstream operational issues.

Validation Across Diverse Operating Models

Testing time logic across diverse operating models to ensure rule consistency and operational stability.

DESIGN PROCESS OVERVIEW

1. Restaurant Research
Layered research to uncover real restaurant workflows
To ensure the scheduling system could support diverse restaurant models, I analyzed 100+ merchant feedback entries, conducted interviews, and reviewed operational data to understand structural workflow differences. This layered approach allowed me to design a solution that scales across shift types and operating hours without increasing staff workload.
1. Customer Research
Efficient validation through market benchmarks
To validate customer-side interactions efficiently, I benchmarked four mature ordering platforms instead of conducting broad user research from scratch. This ensured the design aligned with familiar customer patterns while allowing focus on solving merchant-side complexity.
1. System-Level Research
Map system constraints and time-related dependencies across platforms
To identify structural constraints, I mapped time-related logic across the Merchant Platform, POS, and Online Ordering systems to surface dependencies and timing conflicts. This system-level analysis revealed legacy mechanisms disrupting scheduling consistency, allowing rule alignment before introducing new features.
1. Product Strategy
Drive a high-impact release strategy through business goal breakdown and feasibility alignment
To shape a high-impact MVP, I translated business goals into adoption-focused metrics and aligned scope through feasibility reviews with PM and Engineering. Rather than expanding feature breadth, I prioritized low-friction setup and compatibility with existing workflows to drive GMV growth without increasing operational complexity.
1. Scenario & Edge Case Mapping
Designing cross-role state mapping to eliminate edge cases and operational risk
To prevent cross-role scheduling failures, I mapped full state transitions across customer, merchant, and kitchen workflows to surface hidden dependencies and timing conflicts. This systematic state mapping allowed me to identify edge cases early and eliminate operational risks before implementation.
1. Pilot Test
Validate workflow safety through pilot testing before launch
Before launch, I conducted pilot testing with selected restaurants to validate real-world workflow safety. Rather than focusing on feature feedback, the test evaluated operational stability and cross-role coordination, confirming no critical breakdowns and enabling a low-risk rollout.

1. Restaurant Research
Layered research to uncover real restaurant workflows
To ensure the scheduling system could support diverse restaurant models, I analyzed 100+ merchant feedback entries, conducted interviews, and reviewed operational data to understand structural workflow differences. This layered approach allowed me to design a solution that scales across shift types and operating hours without increasing staff workload.
1. Customer Research
Efficient validation through market benchmarks
To validate customer-side interactions efficiently, I benchmarked four mature ordering platforms instead of conducting broad user research from scratch. This ensured the design aligned with familiar customer patterns while allowing focus on solving merchant-side complexity.
1. System-Level Research
Map system constraints and time-related dependencies across platforms
To identify structural constraints, I mapped time-related logic across the Merchant Platform, POS, and Online Ordering systems to surface dependencies and timing conflicts. This system-level analysis revealed legacy mechanisms disrupting scheduling consistency, allowing rule alignment before introducing new features.
1. Product Strategy
Drive a high-impact release strategy through business goal breakdown and feasibility alignment
To shape a high-impact MVP, I translated business goals into adoption-focused metrics and aligned scope through feasibility reviews with PM and Engineering. Rather than expanding feature breadth, I prioritized low-friction setup and compatibility with existing workflows to drive GMV growth without increasing operational complexity.
1. Scenario & Edge Case Mapping
Designing cross-role state mapping to eliminate edge cases and operational risk
To prevent cross-role scheduling failures, I mapped full state transitions across customer, merchant, and kitchen workflows to surface hidden dependencies and timing conflicts. This systematic state mapping allowed me to identify edge cases early and eliminate operational risks before implementation.
1. Pilot Test
Validate workflow safety through pilot testing before launch
Before launch, I conducted pilot testing with selected restaurants to validate real-world workflow safety. Rather than focusing on feature feedback, the test evaluated operational stability and cross-role coordination, confirming no critical breakdowns and enabling a low-risk rollout.

LEARNING

❶ Evaluate real workflows before adding system rules
I enforced “pickup time must be within business hours” assuming it reflected real restaurant operations. But this rule added days of extra design and engineering work, while restaurants could already self-check and correct their settings.
I learned to only enforce system-level safeguards when mistakes create major, irreversible impact. Real workflows and actual risk must guide whether a rule is necessary.
❷ Question existing structures instead of assuming they work
I initially placed the feature based on the platform’s existing information architecture, assuming users would locate it easily. Pilot testing showed they could not, which led to rework. This taught me to challenge inherited designs and validate user behavior early, especially for core flows that directly affect activation and business outcomes.
❸ Preserve familiarity when improving fast-paced workflows
Restaurant operations move quickly and rely heavily on muscle memory. Even small changes in layout or flow can slow staff down or introduce errors during busy periods. This taught me that clarity is important, but maintaining familiar patterns is equally critical. When designing for fast-paced environments like restaurants, improvements must enhance understanding without disrupting the flow people already depend on.
❶ Evaluate real workflows before adding system rules
I enforced “pickup time must be within business hours” assuming it reflected real restaurant operations. But this rule added days of extra design and engineering work, while restaurants could already self-check and correct their settings.
I learned to only enforce system-level safeguards when mistakes create major, irreversible impact. Real workflows and actual risk must guide whether a rule is necessary.
❷ Question existing structures instead of assuming they work
I initially placed the feature based on the platform’s existing information architecture, assuming users would locate it easily. Pilot testing showed they could not, which led to rework. This taught me to challenge inherited designs and validate user behavior early, especially for core flows that directly affect activation and business outcomes.
❸ Preserve familiarity when improving fast-paced workflows
Restaurant operations move quickly and rely heavily on muscle memory. Even small changes in layout or flow can slow staff down or introduce errors during busy periods. This taught me that clarity is important, but maintaining familiar patterns is equally critical. When designing for fast-paced environments like restaurants, improvements must enhance understanding without disrupting the flow people already depend on.
❶ Evaluate real workflows before adding system rules
I enforced “pickup time must be within business hours” assuming it reflected real restaurant operations. But this rule added days of extra design and engineering work, while restaurants could already self-check and correct their settings.
I learned to only enforce system-level safeguards when mistakes create major, irreversible impact. Real workflows and actual risk must guide whether a rule is necessary.
❷ Question existing structures instead of assuming they work
I initially placed the feature based on the platform’s existing information architecture, assuming users would locate it easily. Pilot testing showed they could not, which led to rework. This taught me to challenge inherited designs and validate user behavior early, especially for core flows that directly affect activation and business outcomes.
❸ Preserve familiarity when improving fast-paced workflows
Restaurant operations move quickly and rely heavily on muscle memory. Even small changes in layout or flow can slow staff down or introduce errors during busy periods. This taught me that clarity is important, but maintaining familiar patterns is equally critical. When designing for fast-paced environments like restaurants, improvements must enhance understanding without disrupting the flow people already depend on.
❶ Evaluate real workflows before adding system rules
I enforced “pickup time must be within business hours” assuming it reflected real restaurant operations. But this rule added days of extra design and engineering work, while restaurants could already self-check and correct their settings.
I learned to only enforce system-level safeguards when mistakes create major, irreversible impact. Real workflows and actual risk must guide whether a rule is necessary.
❷ Question existing structures instead of assuming they work
I initially placed the feature based on the platform’s existing information architecture, assuming users would locate it easily. Pilot testing showed they could not, which led to rework. This taught me to challenge inherited designs and validate user behavior early, especially for core flows that directly affect activation and business outcomes.
❸ Preserve familiarity when improving fast-paced workflows
Restaurant operations move quickly and rely heavily on muscle memory. Even small changes in layout or flow can slow staff down or introduce errors during busy periods. This taught me that clarity is important, but maintaining familiar patterns is equally critical. When designing for fast-paced environments like restaurants, improvements must enhance understanding without disrupting the flow people already depend on.

Scheduled Ordering System

Designed a restaurant scheduling system from zero — solving cross-role workflow failures and driving 53% adoption in the first month. Scaled to 16K+ restaurants and 2M+ customers.

The system didn’t support scheduled orders, leading to manual workarounds, lost sales, and cross-role operational confusion

Designing across 4 roles, 3 platforms, and diverse restaurant business models without disrupting existing operations

Achieved 53% adoption in the first month, driving 1%+ online revenue growth

+53%

+1%

Research across both sides of the system — restaurants and customers — to define what a reliable scheduling system needs to do

Restaurants have diverse workflows, but share the same core need: fewer missed orders, less effort

💰

Protect revenue from scheduling mistakes, off-hour loss

👥

Less manual coordination and follow-ups

🍽️

Prepare orders at the right time

Customers want a clear and low-effort way to schedule orders and pick up on time

Make scheduling availability visible upfront

Make pickup time selection clear and predictable

Designing an end-to-end experience by making deliberate trade-offs across roles, platforms, and mental models

Easy, flexible setup for diverse restaurant workflows

Rely on customers to manually specify pickup times in order notes

Define pickup rules at the system level instead of relying on manual instructions

❌ Navigation placement Idea A: Reusing existing IA with low discoverability and high density

✅ Navigation placement Idea B: A dedicated entry point designed for clarity and future scale

A clear scheduling experience built on a single source of truth

The system only supported ASAP orders, forcing customers write manual pickup notes, creating confusion across workflows.

Pickup options are generated from restaurant-configuration, creating a single, reliable pickup time across workflows.

Conflicting pickup times from system output and manual notes caused confusion.

A single, system-defined pickup time creates clarity, trust, and consistency.

A minimal-disruption workflow for fast order management

Configurable "Print-Later" action to support diverse restaurant workflows and maximize adoption

The order confirmation view wasn’t aligned with staff workflows, slowing down order decisions.

A workflow-aligned confirmation view enables faster, clearer order decisions.

Designed and tested multiple UI concepts to evaluate trade-offs and determine the final direction

Preserved familiar “Accept” behavior and added “Print Later” to support diverse workflows without retraining.

A single source of truth that eliminated conflicting pickup times

Conflicting pickup times from notes and system confused kitchen staff

A single system-defined pickup time provides a clear source of truth for meal prep

Rebuilding the scheduling logic to eliminate conflicts across roles and platforms

Defining pickup time rules to ensure clarity and a single source of truth

① Availability from pickup windows, not calendar dates

② ASAP is always the earliest valid option

③ ASAP time calculated at order placement

Reducing system complexity and engineering overhead by sunsetting legacy auto-cancel logic

Reducing system complexity and engineering overhead by sunsetting legacy auto-cancel logic

Validating the new logic across real operating models

Validation Across Diverse Operating ModelsTesting time logic across diverse operating models to ensure rule consistency and operational stability.

Layered research to uncover real restaurant workflows

Efficient validation through market benchmarks

Map system constraints and time-related dependencies across platforms

Drive a high-impact release strategy through business goal breakdown and feasibility alignment

Scenario & Edge Case Mapping

Designing cross-role state mapping to eliminate edge cases and operational risk

Validate workflow safety through pilot testing before launch

Layered research to uncover real restaurant workflows

Efficient validation through market benchmarks

Map system constraints and time-related dependencies across platforms

Drive a high-impact release strategy through business goal breakdown and feasibility alignment

Scenario & Edge Case Mapping

Designing cross-role state mapping to eliminate edge cases and operational risk

Validate workflow safety through pilot testing before launch

❶ Evaluate real workflows before adding system rules

❷ Question existing structures instead of assuming they work

❸ Preserve familiarity when improving fast-paced workflows

Curiosity & Strategy

Designing real-world impact across UX, systems, and teams.

Curiosity & Strategy

Designing real-world impact across UX, systems, and teams.

Curiosity & Strategy

Designing real-world impact across UX, systems, and teams.

Validation Across Diverse Operating Models

Testing time logic across diverse operating models to ensure rule consistency and operational stability.