🗽 NYC 311 Chat Assistant Project Wiki

# User Needs Template for Chatbot Design

## 1. Information Needs

What the user seeks to know or understand:
* [Primary knowledge gap or question they need answered]
* [Structured guidance or framework to navigate their situation]
* [Communication strategies for stakeholders in their ecosystem]
* [Access to resources, tools, or human connections]
* [Practical, actionable answers to immediate concerns]

**Example Queries:**
* "[Core question about their situation]"
* "[Question about capability/limitation during process]"
* "[Question about consequences or outcomes]"


## 2. Expectations

What the user expects from the chatbot experience:
* [Emotional tone and approach - e.g., empathetic, professional, friendly]
* [Accuracy and credibility of information provided]
* [Level of personalization based on their context]
* [Availability and responsiveness requirements]
* [Escalation path to human assistance when needed]


## 3. Motivations

What drives the user to engage with the chatbot:
* [Desire for control, autonomy, or empowerment]
* [Need for connection, support, or validation]
* [Goal to prepare themselves or others]
* [Search for clarity, certainty, or understanding]


## 4. Demands and Preferences

**Preferences (Nice-to-haves):**
* [Communication style and tone - e.g., conversational, formal, warm]
* [Information presentation format - e.g., simplified, step-by-step, detailed]
* [Interaction style - e.g., ability to ask follow-ups, explore topics]
* [Multimedia support - e.g., visuals, videos, documents]
* [Access channels - e.g., text, voice, mobile, desktop]

**Demands (Must-haves):**
* [Security/privacy requirements]
* [Personalization boundaries - avoiding generic responses]
* [Speed and efficiency expectations]


## 5. Anxieties and Concerns

Emotional and practical barriers the user may experience:
* [Feeling overwhelmed by complexity or volume of information]
* [Fear or worry about outcomes or consequences]
* [Concerns about impact on their ecosystem - relationships, work, life]
* [Uncertainty about where to begin or what's trustworthy]
* [Psychological safety concerns - judgment, appearing uninformed]

# Bot Needs Template for Chatbot Design (Simplified)

## 1. Core Capabilities

What the bot is designed to do:
* [Primary Function Area 1]
   * [Specific capability or action]
   * [Specific capability or action]
* Examples:
   * ✅ DO: [Concrete example of good implementation]
   * ❌ DON'T: [What to avoid]

* [Primary Function Area 2]
   * [Specific capability or action]
   * [Specific capability or action]
* Examples:
   * ✅ DO: [Concrete example of good implementation]
   * ❌ DON'T: [What to avoid]

* [Primary Function Area 3]
   * [Specific capability or action]
* Examples:
   * ✅ DO: [Concrete example of good implementation]
   * ❌ DON'T: [What to avoid]


## 2. Must Ask

Information the bot needs to collect for personalization:
* [Key data point 1]
* [Key data point 2]
* [Key data point 3]

Examples:
* ✅ DO: [How to ask with context/reason]
* ❌ DON'T: [Form-style interrogation or asking without reason]


## 3. Has Already

Information available to the bot from existing systems:
* [Pre-existing data source 1]
* [Pre-existing data source 2]
* [Pre-existing data source 3]

Examples:
* ✅ DO: [How to use existing data transparently]
* ❌ DON'T: [Creepy or outdated data usage]


## 4. Must Provide

Non-negotiable outputs or resources:
* [Core deliverable 1]
* [Core deliverable 2]
* [Core deliverable 3]

Examples:
* ✅ DO: [Actionable, multi-format delivery]
* ❌ DON'T: [Info dump without next steps]


## 5. User Behavior Patterns

Expected user interaction styles:
* [Behavior Type 1]: [Description]
   * ✅ DO: [How bot should respond]
   * ❌ DON'T: [What frustrates this user]

* [Behavior Type 2]: [Description]
   * ✅ DO: [How bot should respond]
   * ❌ DON'T: [What frustrates this user]

* [Behavior Type 3]: [Description]
   * ✅ DO: [How bot should respond]
   * ❌ DON'T: [What frustrates this user]


## 6. Limitations

What the bot cannot or must not do:
* [Hard Boundary 1]
   * ❌ DON'T: [Example of violation]
   * ✅ DO INSTEAD: [Alternative safe response]

* [Hard Boundary 2]
   * ❌ DON'T: [Example of violation]
   * ✅ DO INSTEAD: [Alternative safe response]

* [Hard Boundary 3]
   * ❌ DON'T: [Example of violation]
   * ✅ DO INSTEAD: [Alternative safe response]


## 7. Tone & Approach

Personality and communication style:
[Overall character/role in 1 sentence]

Core tone: [Adjective], [Adjective], [Adjective]

Examples:

[Tone Quality 1]:
* ✅ DO: "[Example phrase or approach]"
* ❌ DON'T: "[Counter-example to avoid]"

[Tone Quality 2]:
* ✅ DO: "[Example phrase or approach]"
* ❌ DON'T: "[Counter-example to avoid]"

[Tone Quality 3]:
* ✅ DO: "[Example phrase or approach]"
* ❌ DON'T: "[Counter-example to avoid]"


## 📋 Quick Reference Card

| Section           | Key Question                        |
|-------------------|-------------------------------------|
| Core Capabilities | What can the bot actually do?       |
| Must Ask          | What info does the bot need?        |
| Has Already       | What does the bot already know?     |
| Must Provide      | What must the bot always deliver?   |
| User Behavior     | How will users actually interact?   |
| Limitations       | What are the hard boundaries?       |
| Tone & Approach   | How should the bot sound?           |


## ✅ Quality Checklist

Before finalizing, ensure:
* Every ✅ DO example uses realistic user language
* Every ❌ DON'T shows a real anti-pattern to avoid
* Examples don't conflict across sections
* Tone examples can be copy-pasted as starter prompts
* Limitations show both boundary AND alternative
* Behaviors reflect real users, not "ideal" users