Search Ad Agent - Technical Overview

⚠️ Important: This use case demonstrates AI-driven search prompts with conversational chat interfaces. As users type, AI-powered branded suggestions appear in the dropdown with contextual prompts. When clicked, the suggestion morphs into a chat interface with the brand's AI assistant - providing an engaging, interactive shopping experience.

📺 Live Demo: See this implementation in action at https://demo-aura.affinity.net/search-ad-agent

Executive Summary

This document outlines how search publishers can integrate AI-driven branded chat experiences into search autocomplete using standard IAB Native Ads 1.2 format. As users type in the search box, branded suggestions with AI-generated prompts appear in the dropdown. When users click a suggestion, the interface seamlessly transitions to a conversational chat with the brand's AI assistant - providing an immersive, interactive shopping experience.

Key Benefits:

• AI-Driven Prompts: Context-aware questions that engage users
• Conversational Interface: Seamless transition to branded chat experience
• Intent Detection: Smart prompts based on user's search query
• Industry Standard: Uses OpenRTB 2.6 + IAB Native Ads 1.2 specification
• Easy Integration: Works with existing ad tech ecosystem (DSPs, SSPs, ad servers)
• Engagement-Based Revenue: Revenue from chat interactions and engagement
• Privacy-First: Contextual targeting without PII requirements
• Immersive Experience: Morphing animation creates seamless transition
• Brand Differentiation: Unique AI assistant experience per brand

Compare With:

• Search Ad Link - Simple click-through behavior without chat interface
• Homepage Tile Agent - Similar chat experience but for browser homepage tiles

Use Case Example

User Action: User types "hotels in paris" in search box

Autocomplete Dropdown Appears:

┌─────────────────────────────────────────────────────────┐
│ Search: hotels in paris                           [🔍] │
├─────────────────────────────────────────────────────────┤
│ AI Suggestions:                              SPONSORED  │
│                                                         │
│ ┌─────────────────────────────────────────────────┐   │
│ │ [🏨] Planning a trip? Ask Booking.com AI    [→]│   │
│ └─────────────────────────────────────────────────┘   │
│                                                         │
│ ┌─────────────────────────────────────────────────┐   │
│ │ [✈️] Need travel help? Ask Expedia AI        [→]│   │
│ └─────────────────────────────────────────────────┘   │
│                                                         │
│ ┌─────────────────────────────────────────────────┐   │
│ │ [🏨] Looking for hotels? Ask Hotels.com AI   [→]│   │
│ └─────────────────────────────────────────────────┘   │
└─────────────────────────────────────────────────────────┘

User Clicks Booking.com Suggestion:

1. Phase 1 (200ms): Brand pill appears with icon and name, search query displayed
2. Phase 2 (200ms): Interface transitions to chat mode, search bar moves to bottom
3. User's search query becomes first message bubble
4. Bot typing indicator appears (800ms)
5. Context-aware welcome message: "Hi there! 👋 I'm your Booking.com shopping assistant. Great choice searching for 'hotels in paris'! ✈️ I can help you find the best travel options through Booking.com. Where would you like to go?"

User Continues Conversation:

┌─────────────────────────────────────────────────────────┐
│ [🏨 Booking.com]                              [← Back] │
├─────────────────────────────────────────────────────────┤
│                                                         │
│  You: hotels in paris                                   │
│                                                         │
│  🏨 Bot: Hi there! 👋 I'm your Booking.com shopping    │
│         assistant. Great choice searching for "hotels   │
│         in paris"! ✈️ I can help you find the best     │
│         travel options...                               │
│                                                         │
│  You: What are the best areas to stay?                  │
│                                                         │
│  🏨 Bot: Thank you for your interest! This is a demo   │
│         showcasing AI-driven search prompts and         │
│         conversational interfaces. In a production      │
│         environment, I would help you browse            │
│         Booking.com's full catalog...                   │
│                                                         │
├─────────────────────────────────────────────────────────┤
│ [🏨] Type a message...                          [Send] │
└─────────────────────────────────────────────────────────┘

Result: Engaging, conversational brand experience with AI assistant

Why AI-Driven Search Prompts with Chat?

1. Enhanced User Engagement

• AI-generated prompts match user intent from search query
• Conversational interface provides immersive brand experience
• Multi-turn conversations increase engagement time
• Context-aware responses feel natural and helpful

2. Intent-Based Targeting

• Detect user intent from search keywords (gifts, travel, deals, etc.)
• Generate contextual prompts that resonate with user needs
• Welcome messages reference the user's actual search query
• Personalized experience without requiring PII

3. Brand Differentiation

• Each brand gets unique AI assistant personality
• Branded colors and icons throughout experience
• Seamless morphing animation creates premium feel
• Memorable interaction increases brand recall

4. Industry Standard Foundation

• Built on OpenRTB 2.6 and IAB Native Ads 1.2 specification
• Works with existing ad tech ecosystem
• Standard bid request/response format
• Easy integration for advertisers

5. Privacy-First Design

• No PII required in bid requests
• Contextual targeting based on search keywords
• GDPR/CCPA compliant by design
• Client-side chat interface (demo)

6. Engagement-Based Monetization

• Track chat session starts and message exchanges
• Measure engagement depth and duration
• Premium CPMs for interactive experiences
• Multiple monetization opportunities per session

Architecture Overview

Key Points:

• Search platform provides the interface - Both autocomplete and chat UI
• Affinity AI Bid API provides branded suggestions - Standard OpenRTB auction
• AI-driven prompts - Generated client-side based on search intent
• Seamless transition - 2-phase morphing animation (~1.2s total)
• Context-aware chat - Welcome message references user's search query
• Engagement tracking - Session starts, messages, duration

IAB Native Ads Asset Mapping

Search Ad Native Assets

Following IAB Native Ads 1.2 specification, search ad agents use these standard asset types:

Required Assets:

IAB Asset Type	ID	Data Type	Purpose
Title	1	-	Brand name
Data (URL)	5	type: 11	Display URL (brand domain)

Optional Assets:

IAB Asset Type	ID	Data Type	Purpose
Image	2	-	Brand icon (48x48)
Data (CTA)	3	type: 12	Call-to-action text
Data (Sponsor)	4	type: 1	Sponsored label

Note: The click URL (link.url) is used for tracking but not for direct navigation in this use case. Instead, clicking opens the chat interface.

Native Request Structure

The native request follows IAB Native Ads 1.2 specification:

{
  "ver": "1.2",
  "assets": [
    {
      "id": 1,
      "required": 1,
      "title": {
        "len": 100
      }
    },
    {
      "id": 2,
      "required": 0,
      "img": {
        "type": 1,
        "w": 48,
        "h": 48
      }
    },
    {
      "id": 3,
      "required": 1,
      "data": {
        "type": 12,
        "len": 50
      }
    },
    {
      "id": 4,
      "required": 1,
      "data": {
        "type": 1,
        "len": 20
      }
    },
    {
      "id": 5,
      "required": 1,
      "data": {
        "type": 11,
        "len": 100
      }
    }
  ]
}

Key Points:

• Asset ID 1: Brand name (required, max 100 chars)
• Asset ID 2: Brand icon (optional, 48x48px)
• Asset ID 3: CTA text (required, max 50 chars)
• Asset ID 4: Sponsored label (required, max 20 chars)
• Asset ID 5: Display URL (required, max 100 chars, data type 11)

OpenRTB Integration

Note: We use standard OpenRTB 2.6 with IAB Native Ads 1.2 for search ad agents.

Bid Request Structure

Your search platform sends standard OpenRTB 2.6 bid requests with native ad placement:

{
  "id": "search-ads-req-1704123456789-abc123",
  "imp": [
    {
      "id": "imp-1",
      "native": {
        "request": "{\"ver\":\"1.2\",\"assets\":[{\"id\":1,\"required\":1,\"title\":{\"len\":100}},{\"id\":2,\"required\":0,\"img\":{\"type\":1,\"w\":48,\"h\":48}},{\"id\":3,\"required\":1,\"data\":{\"type\":12,\"len\":50}},{\"id\":4,\"required\":1,\"data\":{\"type\":1,\"len\":20}},{\"id\":5,\"required\":1,\"data\":{\"type\":11,\"len\":100}}]}"
      },
      "bidfloor": 0.01,
      "bidfloorcur": "USD"
    },
    {
      "id": "imp-2",
      "native": {
        "request": "{\"ver\":\"1.2\",\"assets\":[...]}"
      },
      "bidfloor": 0.01,
      "bidfloorcur": "USD"
    }
    // ... up to 5 impressions total
  ],
  "site": {
    "id": "search-ads.com",
    "domain": "demo-aura.affinity.net",
    "page": "https://demo-aura.affinity.net/search-ad-agent",
    "keywords": "hotels in paris"
  },
  "device": {
    "language": "en",
    "geo": {
      "country": "US"
    }
  }
}

Key Request Components:

• Request ID: Unique identifier (format: "search-ads-req-{timestamp}-{random}")
• Impressions: Array of 5 impression objects (one per ad slot)
• Native Request: JSON-encoded IAB Native Ads 1.2 specification
• Site Keywords: User's search query (REQUIRED for contextual targeting)
• Device: User device and geo information for targeting
• Bid Floor: 0.01 USD minimum CPM

Bid Response Structure

Advertisers respond with standard IAB Native Ads 1.2 format:

{
  "id": "search-ads-req-1704123456789-abc123",
  "seatbid": [
    {
      "seat": "advertiser-network",
      "bid": [
        {
          "id": "booking-com-bid-001",
          "impid": "imp-1",
          "price": 2.5,
          "adm": "{\"native\":{\"ver\":\"1.2\",\"assets\":[{\"id\":1,\"required\":1,\"title\":{\"text\":\"Booking.com\"}},{\"id\":2,\"img\":{\"url\":\"https://cdn.example.com/booking-icon.png\"}},{\"id\":3,\"data\":{\"type\":12,\"value\":\"Ask AI\"}},{\"id\":4,\"data\":{\"type\":1,\"value\":\"Sponsored\"}},{\"id\":5,\"data\":{\"type\":11,\"value\":\"booking.com\"}}],\"link\":{\"url\":\"https://booking.com?ref=aura\"},\"imptrackers\":[\"https://track.example.com/imp?id=123\"]}}",
          "crid": "booking-creative-001",
          "adomain": ["booking.com"]
        }
      ]
    }
  ],
  "cur": "USD"
}

Native Response Breakdown:

The adm field contains a JSON-encoded IAB Native Ads 1.2 response:

• Assets:
  • Asset 1 (Title): Brand name - "Booking.com"
  • Asset 2 (Image): Brand icon URL
  • Asset 3 (Data/CTA): Call-to-action - "Ask AI" (data type 12)
  • Asset 4 (Data/Sponsor): Sponsored label - "Sponsored" (data type 1)
  • Asset 5 (Data/URL): Display URL - "booking.com" (data type 11)
• Link: Click tracking URL (used for analytics, not navigation)
• Impression Trackers: Array of impression tracking URLs

Important: The native response can be in the adm field directly or wrapped in a native object for legacy compatibility. Both formats are supported.

AI Prompt Generation

Intent Detection Patterns

The search platform analyzes the user's search query to detect intent and generate contextual prompts:

Intent Category	Keywords	Example Prompts
Gifts	gift, present	"Need gift ideas?", "Looking for gifts?"
Travel	travel, trip, vacation, flight, hotel	"Planning a trip?", "Need travel help?"
Deals	deal, sale, discount, offer	"Looking for deals?", "Need the best deals?"
Shopping	buy, purchase, shop	"Ready to shop?", "Looking to buy?"
Recommend	best, top, recommend	"Need recommendations?", "Want the best options?"
Budget	cheap, affordable, budget	"Need budget options?", "Looking for affordable?"
Compare	compare, vs, versus	"Need help comparing?", "Want to compare?"
Review	review, rating	"Need product reviews?", "Want expert opinions?"
Electronics	tv, television, electronics, laptop, phone	"Looking for tech?", "Need electronics help?"
Food	snack, food, grocery, popcorn	"Looking for snacks?", "Need food ideas?"
Fashion	clothes, fashion, outfit, shirt, dress	"Need fashion help?", "Looking for style?"
Home	home, furniture, decor, kitchen	"Need home items?", "Looking for decor?"
Books	book, read, kindle	"Looking for books?", "Need reading ideas?"
Toys	toy, game, kids	"Need toys?", "Looking for games?"
Beauty	beauty, makeup, skincare	"Need beauty products?", "Looking for makeup?"
General	No match	"Looking for {query}?", "Need help with {query}?"

Prompt Format

All prompts follow the format: "{question}? Ask {Brand} AI"

Examples:

• "Planning a trip? Ask Booking.com AI"
• "Looking for deals? Ask Amazon AI"
• "Need gift ideas? Ask Etsy AI"

Diversity Rules

• Each ad receives a unique prompt when possible
• System avoids duplicate questions in the same dropdown
• Falls back to variations if needed for multiple ads

Chat Transition Experience

2-Phase Optimized Animation

The transition from search to chat is optimized for immediate feedback and smooth user experience:

Phase 1: Immediate Display (200ms)

• Immediately show context pill with brand icon and name
• Display user's search query in search box (no typing animation)
• Apply brand color to border
• Provides instant visual feedback

Phase 2: Transition to Chat (200ms)

• Transition to chat mode
• Search bar moves to bottom
• User's search query becomes first message bubble
• Show bot typing indicator (800ms)
• Display context-aware welcome message

Total Animation Time: ~1.2 seconds (400ms animation + 800ms typing indicator)

Key UX Improvements

• No slow typing animations - Instant display of search query
• Immediate visual feedback - Brand pill appears immediately on click
• User's actual search query - Becomes the first message (not "Hello!")
• Much faster transition - ~1.2s vs traditional ~2-3s animations
• Seamless morphing - Search bar smoothly transitions to chat input

Context-Aware Welcome Messages

Intent-Based Greetings

The bot's welcome message is dynamically generated based on the user's original search query, creating a natural conversation flow:

Format: "Hi there! 👋 I'm your {Brand} shopping assistant. {Contextual greeting with query reference}"

Examples by Intent:

1. Gifts Intent (keywords: gift, present)

   • "I see you're searching for '{query}'! I can help you find the perfect gift at {Brand}. What's the occasion?"

2. Travel Intent (keywords: travel, trip, vacation, flight, hotel)

   • "Great choice searching for '{query}'! ✈️ I can help you find the best travel options through {Brand}. Where would you like to go?"

3. Deals Intent (keywords: deal, sale, discount, offer)

   • "I found your search for '{query}'! You're in the right place for deals. {Brand} has amazing offers. Let me help you find the best ones!"

4. Electronics Intent (keywords: tv, television, electronics, laptop, phone)

   • "Searching for '{query}'? {Brand} has great tech options. Let me help you find the perfect device!"

5. General Fallback (no match)

   • "I see you're searching for '{query}'. I can help you find products, compare prices, and answer questions about {Brand}. What would you like to know?"

Benefits

• Personalized experience - References user's actual search query
• Natural conversation - Feels like the bot understood the user's intent
• Works for unlimited brands - No per-brand configuration needed
• Relevant from first message - Maintains context throughout conversation
• Scalable - Single intent detection system for all brands

Chat Interaction Flow

Message Exchange

User Messages:

• User types message in chat input
• Message text is trimmed and validated (not empty, max 2000 chars)
• User message bubble appears immediately
• Message includes timestamp and unique ID

Bot Responses (Demo):

• Generic demo response for all user messages after welcome
• Explains this is a demo showcasing AI-driven search prompts
• Mentions production capabilities (browse catalog, answer questions, assist shopping)
• Encourages exploring other brands
• Brand name dynamically injected

Demo Response Format:

"Thank you for your interest! This is a demo showcasing AI-driven search prompts and conversational interfaces. In a production environment, I would help you browse {Brand}'s full catalog, answer specific questions, and assist with your shopping needs. Feel free to explore other brands by returning to search! 🔍"

Production Integration

In a production environment, the chat interface would integrate with:

• Brand's AI/LLM backend - For intelligent responses
• Product catalog APIs - For real product recommendations
• Inventory systems - For availability and pricing
• Order management - For purchase assistance
• Customer service - For support queries

Ad Display Integration

Parsing Native Ad Response

Your search platform extracts ad information from the IAB Native Ads 1.2 assets:

1. Parse the JSON-encoded adm field from the bid response
2. Handle both wrapped ({native: {...}}) and unwrapped formats
3. Extract native assets by ID:
   • Asset ID 1 (Title) → Brand name
   • Asset ID 2 (Image) → Brand icon URL (optional)
   • Asset ID 3 (Data type 12) → CTA text
   • Asset ID 4 (Data type 1) → Sponsored label
   • Asset ID 5 (Data type 11) → Display URL
   • Link URL → Click tracking URL
   • Impression trackers → Array of tracking URLs
4. Validate all URLs are HTTPS
5. Validate required fields (brand name, click URL)
6. Generate AI-driven prompt based on search query intent
7. Assign brand color from predefined map (or fallback to Google Blue)
8. Render ad in suggestions dropdown

Rendering Search Suggestion

Display the search ad suggestion in the autocomplete dropdown:

• Brand Icon: From Asset ID 2 or first letter of brand name
• AI Prompt: Generated based on search intent + brand name
• Arrow Icon: Visual indicator for clickable action
• Hover State: Background color change on hover
• Keyboard Navigation: Support arrow keys and enter

Visual Design:

┌─────────────────────────────────────────────────┐
│ [🏨] Planning a trip? Ask Booking.com AI    [→]│
└─────────────────────────────────────────────────┘

Loading State:

• "Loading suggestions..." message while fetching ads
• Smooth fade-in when ads load
• Graceful handling if no ads returned

Click Behavior

When user clicks the AI prompt suggestion:

1. Fire Impression Tracker: Send beacon to impression tracking URLs
2. Start Chat Session: Generate unique session ID
3. Trigger Transition: Start 2-phase morphing animation
4. Initialize Chat: Set up chat state with active brand
5. Display Welcome: Show context-aware welcome message with search query
6. Track Session Start: Log chat session initiation for analytics

Implementation:

function handleSuggestionClick(ad: SearchAd, searchQuery: string) {
  // Fire impression trackers
  if (ad.impressionTrackers) {
    ad.impressionTrackers.forEach(url => {
      if (navigator.sendBeacon) {
        navigator.sendBeacon(url)
      }
    })
  }

  // Track suggestion view
  trackingService.trackSuggestionView(ad, searchQuery)

  // Start chat session
  trackingService.trackChatSessionStart(ad)

  // Trigger transition with search query context
  startChatTransition(ad, searchQuery)
}

Tracking Implementation

Track user engagement throughout the experience:

Suggestion View Tracking:

• Fired when AI prompt appears in dropdown
• Includes: ad ID, brand, prompt, search query, timestamp
• Console logging only (demo)

Chat Session Tracking:

• Fired when user enters chat mode
• Generates unique session ID: "session-{timestamp}-{random}"
• Includes: ad ID, brand, session ID, timestamp

Message Tracking:

• Fired for both user and bot messages
• Logs message type and length (not content for privacy)
• Includes: session ID, message type, length, timestamp

Implementation Details:

• Uses navigator.sendBeacon() for reliable tracking
• All tracking URLs validated as HTTPS
• Graceful degradation if tracking fails
• Privacy-conscious: no message content logged

User Experience Flow

1. Search and Suggestions Display

• User types in search box (minimum 3 characters)
• 300ms debounce delay before API call
• AI-driven suggestions appear in dropdown
• Each suggestion shows: brand icon, AI prompt, arrow icon
• "SPONSORED" label at top of dropdown
• Suggestions update as user continues typing
• Entire suggestion is clickable

2. Suggestion Click → Chat Transition

• User clicks AI prompt suggestion
• Phase 1 (200ms): Brand pill appears, search query displayed
• Phase 2 (200ms): Interface morphs to chat mode
• Search bar moves to bottom
• User's search query becomes first message bubble
• Bot typing indicator appears (800ms)
• Context-aware welcome message displays

3. Chat Conversation

• User types messages in chat input
• User messages appear immediately
• Bot typing indicator shows before responses
• Bot responses appear with brand avatar
• Smooth auto-scroll to latest message
• Back button available to return to search

4. Return to Search

• User clicks back button
• Immediate transition back to search mode
• Chat state cleared
• Session tracking reset
• Search input ready for new query

5. Error Handling

• Graceful degradation if no suggestions returned
• Loading state shown while fetching
• No suggestions shown for queries <3 characters
• Network errors return empty array (no user-facing error)
• Invalid ads skipped during parsing

Benefits for Search Publishers

Revenue Generation

• Engagement-Based: Revenue tied to chat interactions and session depth
• High Intent: User typing indicates active search intent
• Premium Placements: AI-driven prompts command premium rates
• Multiple Touchpoints: Track views, clicks, sessions, messages
• Brand Differentiation: Unique experiences increase advertiser demand

User Experience

• Engaging Interface: Conversational AI creates immersive experience
• Intent-Aware: Prompts match user's search intent
• Fast: Instant suggestions with 300ms debounce
• Smooth Transitions: Optimized 2-phase animation (~1.2s)
• Transparent: Clear "SPONSORED" label on suggestions
• Contextual: Welcome messages reference user's search query

Technical Advantages

• Industry Standard: OpenRTB 2.6 + IAB Native Ads 1.2
• Easy Integration: Works with existing ad tech infrastructure
• Client-Side AI: Prompt generation happens in browser (no backend needed)
• Scalable: Handle millions of autocomplete requests per day
• Flexible: Works for unlimited brands without per-brand configuration
• Broad Support: Compatible with all major ad platforms

Analytics & Optimization

• Rich Metrics: Track views, clicks, sessions, messages, duration
• Intent Analysis: Understand which intents drive engagement
• A/B Testing: Test different prompt styles and transition animations
• Performance Data: Optimize based on engagement rates
• Revenue Reporting: Standard programmatic advertising metrics

Implementation Considerations

Quality Control

• Asset Validation: Ensure all required native assets are present (brand name, display URL)
• HTTPS Validation: All URLs (icons, tracking) must be HTTPS
• XSS Prevention: Sanitize all text content before rendering
• Flexible Parsing: Handle both wrapped and unwrapped native response formats
• Performance: Target <50ms per ad parsing
• URL Validation: Ensure URLs are valid before using

Privacy & Compliance

• No PII Required: Contextual targeting only
• GDPR Compliant: User consent management
• CCPA Compliant: Privacy string support
• IAB Guidelines: Follow IAB Native Ads disclosure requirements
• Clear Labeling: Mark ads with "SPONSORED" label for transparency
• Privacy-Conscious Tracking: Log message length, not content

Performance

• Low Latency: 5000ms timeout for bid request
• Debouncing: 300ms debounce delay to reduce API calls
• Async Tracking: Fire tracking asynchronously with sendBeacon
• Graceful Degradation: Handle no-bid responses gracefully
• Request Cancellation: Cancel in-flight requests on new search
• Performance Targets:
  • Bid request construction: <10ms
  • Bid API response: <5000ms
  • Suggestion parsing: <50ms per suggestion
  • Prompt generation: <5ms per prompt
  • Dropdown display: <100ms after typing stops
  • Transition animation: ~1.2s total
  • Chat message rendering: <50ms

Security

• XSS Prevention: Sanitize all text content
• HTTPS Only: Reject non-HTTPS URLs
• Content Security: Validate all external resources
• Input Validation: Validate user messages (max length, not empty)

Demo Implementation

A complete working demo is available at https://demo-aura.affinity.net/search-ad-agent:

Features:

• Real Affinity AI Bid API integration (no mock data)
• IAB Native Ads 1.2 format parsing
• AI-driven prompt generation (client-side)
• Intent detection from search queries
• 2-phase optimized morphing animation
• Context-aware welcome messages
• Generic demo bot responses
• 300ms debounce for API calls
• Keyboard navigation (arrow keys, enter, escape)
• Responsive design (desktop and mobile)
• Session tracking and analytics

Demo Behavior:

• Type minimum 3 characters to trigger suggestions
• AI-driven prompts appear in dropdown
• Click suggestion to enter chat mode
• Seamless morphing animation (~1.2s)
• Context-aware welcome message references your search
• Generic demo responses for subsequent messages
• Back button returns to search

Real Implementation:

• Affinity AI Bid API integration
• OpenRTB 2.6 protocol
• IAB Native Ads 1.2 parsing
• Client-side AI prompt generation
• Intent detection algorithms
• Session tracking with sendBeacon
• Branded suggestions from real advertisers

Related Use Cases

• Search Ad Link - Simple click-through search ads without chat interface
• Homepage Tile Agent - Similar AI chat experience but for browser homepage tiles
• AI Search SEM Text Ads - Text ads integrated with organic search results
• AI Search Product Ads - Visual product cards with images

Related Documentation

OpenRTB Components

• Request Format - OpenRTB 2.6 bid request format
• Response Format - OpenRTB 2.6 bid response format
• Native Object - Native ad request structure

External Standards

• OpenRTB 2.6 Specification - IAB Tech Lab
• IAB Native Ads 1.2 - IAB Tech Lab

Version History

• v1.0 (2026-01-09): Initial use case guide
  • AI-driven search prompts with intent detection
  • Conversational chat interface with 2-phase transition
  • Context-aware welcome messages
  • IAB Native Ads 1.2 specification
  • Demo implementation details

Conclusion

AI-driven search prompts with conversational chat interfaces enable search publishers to create engaging, interactive brand experiences that go beyond traditional click-through advertising. This approach ensures:

• Enhanced Engagement: Conversational AI creates immersive experiences
• Intent-Aware: Prompts match user's search intent automatically
• Industry Standard: OpenRTB 2.6 + IAB Native Ads 1.2
• Easy Integration: Works with all major ad platforms
• Smooth UX: Optimized 2-phase transition (~1.2s)
• Context-Aware: Welcome messages reference user's search query
• Privacy-First: Contextual targeting without PII
• Scalable: Works for unlimited brands without per-brand configuration
• Premium Revenue: Engagement-based monetization opportunities

By using standard IAB Native Ads 1.2 with client-side AI prompt generation, you can create premium advertising experiences that engage users through natural conversation while maintaining industry compatibility and privacy standards.