Conditional Responses & User State

Conditional responses allow an Aimogen chatbot to behave differently depending on who the user is, what they’ve already done, and the current state of the conversation. Instead of responding the same way to everyone, the chatbot can adapt deterministically based on user state and context.

This is not AI “intuition”. It is rule-based logic layered on top of AI chat.

---

What Conditional Responses Are #

A conditional response is a response or action that is executed only if specific conditions are met.

Conditions can be based on:

• user state
• conversation state
• environment context
• previous actions or inputs

If conditions are not met, the response does not trigger.

---

User State: What It Means #

User state refers to known attributes about the user at the time of the conversation.

Common user state signals include:

• logged-in vs guest
• WordPress user role
• known user ID
• previously collected lead data
• consent accepted or not
• returning vs first-time visitor

User state is evaluated at runtime and can change during a conversation.

---

Conversation State #

Conversation state refers to what has already happened in the current chat.

Examples:

• conversation just started
• user already provided an email
• user already accepted GDPR consent
• user already saw a welcome message
• a workflow step has completed

Conversation state allows the chatbot to avoid repeating itself and to move forward logically.

---

How Conditional Responses Are Used #

Conditional logic is typically used to:

• show different welcome messages
• skip lead collection for logged-in users
• request missing information only once
• escalate conversations based on prior answers
• prevent repeating compliance messages
• customize responses for customers vs guests

This makes conversations feel intentional, not scripted.

---

Conditions vs Triggers #

Conditions are often evaluated inside triggers or workflows.

• Trigger → defines when evaluation happens
• Condition → defines whether an action runs

A trigger can fire, but the action may still be blocked if conditions are not satisfied.

---

Examples of Conditional Logic #

Logged-In vs Guest #

• if user is logged in → skip email request
• if user is guest → ask for contact details

Role-Based Behavior #

• admins get technical responses
• customers get support-focused replies
• guests get onboarding explanations

Consent State #

• if consent not accepted → show consent message
• if consent accepted → allow normal chat

Lead Data State #

• if email already collected → do not ask again
• if phone missing → request phone number

---

Conditional Responses vs AI Behavior #

Conditional responses:

• are deterministic
• always behave the same way
• do not hallucinate
• do not rely on model interpretation

AI responses:

• are probabilistic
• may vary wording
• may interpret intent differently

Conditions are used to enforce rules AI should not be trusted to enforce alone.

---

Scope and Persistence #

Conditions can apply:

• once per conversation
• throughout the conversation
• only at conversation start
• only after a specific step

They do not persist across chatbots and do not automatically persist across sessions unless persistence is enabled.

---

Frontend vs Backend Context #

On the frontend:

• user state includes roles, login status, cookies
• guest vs user distinction matters

In the backend (Playground):

• user state is always admin
• most frontend conditions are irrelevant
• useful mainly for logic testing

---

What Conditional Responses Do Not Do #

They do not:

• change personas
• train the AI
• modify models
• override placement rules
• bypass consent logic
• infer unknown user data

They operate only on known state.

---

Common Mistakes #

• assuming AI “knows” user state without conditions
• repeating lead requests because state is not tracked
• overlapping conditions that conflict
• forgetting guest vs logged-in differences
• mixing conditional logic into persona prompts

Conditional logic belongs in rules, not in prompts.

---

Best Practices #

• keep conditions explicit
• track state changes clearly
• avoid deep condition nesting
• document why a condition exists
• combine with hardcoded workflows for clarity
• let AI handle language, not rules

Structure first, creativity second.

---

Summary #

Conditional Responses allow Aimogen chatbots to react based on user state and conversation state, ensuring the right message is shown to the right user at the right time. They provide deterministic control over chatbot behavior, preventing repetition, enforcing rules, and enabling personalized flows without relying on AI guesswork. Used correctly, they turn chatbots from generic responders into context-aware assistants.