PRD: Guest Personalization

Product Principle: Outcome Over Features

The requirements, ideas, examples, and proposed features in this PRD are not mandatory implementation instructions.

They should be treated as hypotheses or possible solutions to achieve the product goals. Some ideas may be incorrect, incomplete, too complex, or less effective than alternative approaches.

The most important requirement is to improve the actual goal metrics, such as engagement, CTR, booking conversion, repeat booking, coverage, and recommendation relevance.

The team should not implement a feature only because it is written in the PRD. Each feature should be evaluated based on expected impact, feasibility, data availability, user value, and measurable contribution to the success metrics.

If there is a simpler, faster, or more effective solution that can better improve the goal metrics, the team should propose and prioritize that solution instead.

1. Overview

We are building a dynamic personalization system for guest users who are not logged in.

The system will personalize homepage sections, section titles, restaurant recommendations, destination recommendations, and content ordering based on anonymous user signals such as country, city, device, language, browsing behavior, and cookies.

The goal is to make the guest experience feel relevant even before the user creates an account or logs in.

Example: An Indonesian guest browsing Bangkok on iOS may see sections such as:

Popular with Indonesian Travelers
Recommended for You
Halal-Friendly Favorites
Trending This Week in Bangkok

2. Goals

Primary Goals

Increase engagement from guest users.
Improve click-through rate on personalized sections.
Help anonymous users discover relevant restaurants, cities, and experiences.
Support personalization without requiring login.
Enable dynamic section ordering and section title changes based on user context.
Support scalable guest segmentation across countries, cities, devices, and browsing behavior.

Business Goals

Increase booking conversion from guest traffic.
Improve SEO discovery through city-specific and segment-relevant landing pages.
Cross-sell restaurants, destinations, and experiences.
Support campaign-based targeting for anonymous users.

3. User Type

Guest users are users who are not logged in.

Guest personalization should use anonymous signals only.

Examples of guest user signals:

IP country
Browsing city
Device type
Browser language
Cookie/session history
Viewed restaurants
Viewed destinations
Referral source
Current location if available
Country-to-city travel pattern

4. Scope

In Scope

Personalized homepage sections for guest users.
Dynamic section titles.
Dynamic section ordering.
Dynamic restaurant and destination recommendations.
Country-based personalization.
City-based personalization.
Device-based personalization.
Language-based personalization.
Anonymous collaborative filtering.
Manual segmentation from CMS.
Automatic segmentation using rules and machine learning.
Admin controls to enable, disable, and prioritize guest sections.
Internal links and redirects to city-specific landing pages.

Out of Scope

Personalization using member-only data such as booking history, saved restaurants, loyalty status, or profile preferences.
Member-specific rebooking recommendations.
Logged-in user preference management.
Loyalty-based targeting.

5. Key Guest Personalization Features

5.1 Country-Based Recommendations

The system should personalize content based on the guest user’s detected country.

Examples:

Indonesian guest browsing Bangkok -> “Popular with Indonesian Travelers”
Korean guest browsing Bangkok -> “Korean-Loved in Bangkok”
Singapore guest browsing Pattaya -> “Popular with Singapore Travelers”

5.2 City-Specific Landing and Redirects

The system should support city-specific landing pages and internal linking.

Examples:

Bangkok landing page
Singapore landing page
Pattaya landing page

Guest traffic should be redirected or guided to the most relevant city page based on:

IP location
Browsing history
Search intent
Referral source
Selected destination

5.3 Dynamic Homepage Sections

The system should dynamically change:

Section titles
Section order
Section content
Restaurant cards
Destination cards
Promotional modules

Examples:

“Recommended for You”
“Popular in Bangkok”
“Popular with Indonesian Travelers”
“Halal-Friendly Favorites”
“Trending This Week”
“Near You”
“New & Worth Trying”

5.4 Anonymous Recommendation Engine

For guest users, the recommendation engine should use anonymous signals.

Inputs may include:

IP country
Current city being browsed
Device type
Browser language
Cookie/session behavior
Viewed restaurants
Viewed destinations
Co-visitation behavior
Popular items from similar anonymous users

Recommendation methods:

Collaborative filtering
Co-visitation recommendations
Trending content
Country-to-city popularity
Device-based behavior patterns
Session-based recommendations

Example: If many Indonesian iOS guests browsing Bangkok also view Savoey, After You, and Greyhound Cafe, those restaurants may appear in the “Recommended for You” section.

5.5 Manual Guest Segmentation

Admins should be able to create manual guest segments in CMS.

Example segments:

indonesian_guests_browsing_bangkok
korean_travelers_browsing_bangkok
singapore_guests_browsing_pattaya
ios_guests_in_thailand
chinese_new_year_guests
campaign_bangkok_anniversary_guest_audience

Manual segmentation rules may include:

Country
City
Device
Language
Login status = guest
Campaign eligibility
Promo eligibility
Referral source

5.6 Automatic Guest Segmentation

The system should automatically detect guest segments using rules and machine learning.

Example automatic segments:

traveler_from_KR
traveler_from_ID
ios_user
android_user
interested_in_japanese
interested_in_buffet
interested_in_halal
browsing_bangkok
browsing_singapore
browsing_pattaya

Example rule: If IP country = Korea and browsing city = Bangkok and user is not logged in, assign segment = traveler_from_KR.

Example use: Show “Korean-Loved in Bangkok” and “Hotpot & K-BBQ Right Now.”

6. Guest Segmentation Inputs

The system should support the following guest segmentation inputs:

Location Inputs

IP country
IP city
Selected browsing city
Geo lookup
Latitude/longitude if available
Country-to-city mismatch

Example: IP country = Korea, browsing city = Bangkok -> traveler_from_KR.

Device Inputs

iOS
Android
Desktop
Tablet
Browser type

Example: Device = iOS -> segment = ios_user.

Language and Region Inputs

Browser language
Region setting
Country detected from IP
Country selected by user

Example: Browser language = Korean and browsing city = Bangkok -> prioritize Korean-friendly content.

Behavioral Inputs

Restaurants viewed in current session
Destination pages viewed
Search keywords
Clicked sections
Clicked restaurant cards
Time spent on pages
Cookie/session history
Referral campaign

Example: If a guest views multiple Japanese restaurants, the system may show “Japanese Picks in Bangkok.”

7. Guest Segmentation Outputs

Guest segmentation should produce:

Personalized section list.
Personalized section titles.
Personalized restaurant recommendations.
Personalized destination recommendations.
Personalized section order.
Personalized landing page links.
Campaign-specific sections.
Country-specific content groups.
City-specific content groups.

8. CMS and Admin Requirements

Admins should be able to:

Create guest segments.
Edit guest segment rules.
Enable or disable guest segments.
Assign sections to guest segments.
Configure section priority.
Configure fixed and dynamic sections.
Configure fixed and dynamic section ordering.
Preview homepage output by segment.
Set campaign start and end dates.
Set promo eligibility.
Monitor section performance.

9. Section Configuration Rules

Each section should support the following configuration:

Section Type

Fixed section
Dynamic section

Fixed section: The section must always appear for the selected segment.

Dynamic section: The section may be replaced by another section depending on ranking logic.

Section Order

Fixed order
Dynamic order

Fixed order: The section must remain in the configured position.

Dynamic order: The section position may change depending on relevance score.

Example

For guest users:

“Recommended for You” is a fixed section.
It appears in position 1.
“Popular with Indonesian Travelers” is a dynamic section.
It may appear in positions 2 to 4 depending on relevance.
“Trending This Week in Bangkok” is dynamic.
It may move up if engagement velocity is high.

10. Recommendation Ranking Logic

Guest recommendation ranking should consider:

Country relevance
City relevance
Device relevance
Language relevance
Session behavior
Popularity among similar anonymous users
Restaurant availability
Promotional relevance
Trending velocity
Freshness
Distance, if location is available
Business rules from CMS

Example ranking formula:

Guest Recommendation Score =
Country Match Score
+ City Match Score
+ Device/Language Score
+ Session Behavior Score
+ Collaborative Filtering Score
+ Trending Score
+ Promo Score
+ Availability Score
+ CMS Priority Score

11. Example Guest Scenario

Scenario: Anonymous Guest

User context:

IP country: Indonesia
Device: iOS
Logged-in status: Guest
Browsing city: Bangkok

Recommended homepage sections:

Limited-Time Deals in Bangkok
- Copper Beyond Buffet
- Goji Kitchen
- Kub Kao’ Kub Pla
- Bar-B-Q Plaza
Recommended for You
- Savoey
- After You Dessert
- Greyhound Cafe
- Somtum Der
Popular with Indonesian Travelers
- Goji Kitchen
- Audrey Cafe
- Savoey
- Bar-B-Q Plaza
- After You
Halal-Friendly Favorites
- Yana Restaurant
- Usman Thai Muslim
- Saman Islam
- Santan
Trending This Week in Bangkok
- Pe Aor Tom Yum
- Nara
- Charmgang
- Khao
Near You in Bangkok
- Thonglor Picks
- Asok & Phrom Phong
- Siam & Chidlom
New & Worth Trying
- Nusara
- Haoma
- Elena
- Sushi Masa

guest
traveler_from_ID
ios_user
browsing_bangkok
interested_in_halal

IP address
Geo lookup
Device metadata
Browser language
Cookie/session ID
Viewed restaurants
Viewed cities
Click events
Search events
Promo data
Restaurant metadata
Availability data
CMS section configuration

System Components

The system should include:

Segmentation Engine
- Assigns guest segments using rules and ML.
Recommendation Engine
- Generates restaurant and destination recommendations.
Ranking Engine
- Scores and orders sections and cards.
CMS
- Manages sections, rules, priorities, and campaigns.
Tracking System
- Collects guest behavior events.
API Layer
- Serves personalized homepage content.

Performance Requirements

API response time should be less than 200ms.
System should support 1,000+ dynamic segments.
System should support real-time or near-real-time recommendation updates.
System should support high traffic from anonymous users.
Recommendation logic should degrade gracefully if some signals are missing.

14. Success Metrics

Engagement Metrics

Increase CTR on personalized guest sections by 20%.
Increase guest session time.
Increase number of restaurant card clicks.
Increase number of city page clicks.

Conversion Metrics

Increase guest-to-booking conversion rate by 15%.
Increase guest-to-member signup conversion.
Increase booking conversion from personalized sections.

Coverage Metrics

90% of guest users receive at least one personalized section.
90% of guest sessions receive country or city-based personalization when enough signals are available.

System Metrics

API response time below 200ms.
Support 1,000+ guest segments.
Recommendation fallback available when signals are missing.
CMS rules applied correctly.

15. Fallback Logic

If guest signals are limited, the system should fall back in this order:

Current browsing city.
IP country.
Browser language.
Device type.
Trending in selected city.
Global popular restaurants.
Editor’s picks.

Example: If the system only knows the guest is browsing Bangkok, show:

Popular in Bangkok
Trending This Week in Bangkok
Editor’s Picks in Bangkok

16. Open Questions

What anonymous signals are currently available from tracking?
How long should guest cookie history be stored?
Should guest segments be persisted across sessions?
Which sections must always be fixed for guest users?
What is the minimum data required before showing “Recommended for You”?
How should SEO pages differ from personalized homepage sections?

Keyboard shortcuts

HungryHub Knowledge Base