RNPL User Behavior Analysis

RNPL User Behavior Analysis

Analysis Date: February 2, 2026 Data Period: May 2025 - January 2026 Total RNPL Users Analyzed: 7,812 unique users (complete coverage)

Data Source Update

Revised Analysis - Multi-Channel Update

This analysis was revised on February 2, 2026 to use more complete historical data sources:

• sadb_phone_get_logs (77M rows, Jun 2021 - present) - for phone contact behavior
• sadb_whatsapp_communication_flow (5.2M rows, Dec 2024 - present) - for WhatsApp engagement
• sadb_user_listing_views (1.26B rows, Feb 2022 - present) - for viewing behavior

Previous analysis used only phone contacts, which significantly underestimated HESITANT user engagement. The addition of WhatsApp data reveals a critical channel preference insight that changes our understanding of this segment.

Executive Summary

This analysis examines user behavior patterns for Ejari RNPL (Rent Now, Pay Later) applicants to understand:

1. Pre-application engagement levels (phone contacts as high-intent signal)
2. Differences between rejection reason segments
3. Post-rejection retention patterns
4. Actionable user classification framework

Key Finding: HESITANT Users Are Channel-Selective, Not Disengaged

Critical Insight

HESITANT users are NOT window shoppers - they show dramatically different channel preferences:

• Phone contacts: Only 9.9% engaged
• WhatsApp contacts: 54.5% engaged (5.5x higher!)

This suggests HESITANT users actively search but prefer WhatsApp over phone calls. Previous phone-only analysis significantly underestimated their engagement.

Multi-Channel Engagement by Segment

Segment	Users	Phone %	WhatsApp %	Either (est)	Interpretation
HESITANT	1,601	9.9%	54.5%	~55%	Channel-selective - prefer WhatsApp
UNEMPLOYED	1,735	60.4%	55.7%	~75%	Multi-channel active searchers
DBR	1,268	56.2%	58.9%	~75%	Engaged across all channels
LOW_SIMAH_SCORE	1,014	65.8%	54.7%	~80%	Highest overall engagement
ACCEPTED	90	53.6%	31.1%	~60%	Phone-preferred (traditional approach)

All segments show substantial engagement when considering both channels. The key differentiator is channel preference, not intent level.

Status Distribution

Status	Count	Percentage
REJECTED	5,620	71.9%
NO_STATUS (early applications)	1,941	24.8%
CANCELLED	152	1.9%
ACCEPTED	69	0.9%
UNDER_REVIEW	30	0.4%

Rejection Reason Distribution

Reason	Count	% of Rejected	Description
UNEMPLOYED	1,452	25.8%	No verifiable employment
HESITANT	1,270	22.6%	User withdrew during process
DBR	860	15.3%	Debt-to-Burden Ratio too high
LOW_SIMAH_SCORE	810	14.4%	Low credit score
UNRESPONSIVE	606	10.8%	User stopped responding
UNSPECIFIED	174	3.1%	Reason not documented
PAID_FULL	90	1.6%	User paid full rent instead
REFUSED_DOWN_PAYMENT	57	1.0%	Declined down payment terms
LANDLORD_REJECTION	48	0.9%	Property owner declined
Other	253	4.5%	Various other reasons

Pre-Application Behavior Analysis

Phone Contact Metrics by Rejection Reason

Phone contacts (revealing landlord phone numbers) represent high-intent engagement - users who are actively pursuing rental opportunities.

Rejection Reason	Total Users	Users w/ Contacts	% w/ Contacts	Total Contacts	Avg	Median	P90
HESITANT	1,270	126	9.9%	1,716	13.6	8	34
UNEMPLOYED	1,452	877	60.4%	17,228	19.6	9	45
DBR	860	483	56.2%	8,351	17.3	8	43
LOW_SIMAH_SCORE	810	533	65.8%	9,570	18.0	9	43

Phone Contact Insights

1. HESITANT users avoid phone calls:

   • Only 9.9% made any phone contact vs 56-66% for other segments
   • This does NOT mean they are disengaged (see WhatsApp data below)

2. Financial constraint segments show high phone engagement:

   • 56-66% contact rate indicates serious rental intent via traditional channels
   • These users are comfortable with direct phone contact

3. LOW_SIMAH_SCORE users are most phone-engaged:

   • Highest phone contact rate at 65.8%
   • Comfortable with traditional communication methods

WhatsApp Engagement Analysis

WhatsApp data from sadb_whatsapp_communication_flow (5.2M rows, Dec 2024 - present) reveals critical channel preference differences.

WhatsApp Contact Metrics by Segment

Segment	Total Users	Users w/ WhatsApp	% w/ WhatsApp	Total Contacts	Avg Contacts
HESITANT	1,601	872	54.5%	9,461	10.8
UNEMPLOYED	1,735	966	55.7%	10,874	11.3
DBR	1,268	747	58.9%	9,378	12.6
LOW_SIMAH_SCORE	1,014	555	54.7%	5,847	10.5
ACCEPTED	90	28	31.1%	332	11.9

Channel Preference Insights

1. HESITANT users strongly prefer WhatsApp:

   • WhatsApp: 54.5% vs Phone: 9.9% = 5.5x preference for WhatsApp
   • This explains the previously observed “low engagement” - it was channel-specific, not intent-specific
   • These users ARE actively searching, they just avoid phone calls

2. DBR users are most WhatsApp-engaged:

   • Highest WhatsApp rate at 58.9%
   • Combined with 56.2% phone rate = highly multi-channel active

3. ACCEPTED users prefer phone over WhatsApp:

   • Phone: 53.6% vs WhatsApp: 31.1%
   • Traditional communication approach correlates with approval
   • May indicate more established/formal rental search behavior

4. All segments show ~55% WhatsApp engagement:

   • WhatsApp is a universal channel across all segments
   • The differentiator is phone call willingness, not WhatsApp use

Post-Application Retention

Activity After Rejection (Jan 2026)

Segment	Users with Jan Activity	Avg Contacts in Jan
HESITANT	24 users	5.2
UNEMPLOYED	220 users	9.1
DBR	111 users	7.5
LOW_SIMAH_SCORE	121 users	7.8

Retention Insights

• HESITANT users disengage rapidly - only 24 continued activity post-rejection
• UNEMPLOYED users show highest persistence - 220 users still actively searching
• Financial constraint segments (DBR, LOW_SIMAH) maintain moderate engagement
• Rejected users with employment/credit barriers are still in the rental market and represent re-engagement opportunities

Behavioral Profiles

1. Channel-Selective Searchers (HESITANT) - REVISED

• 9.9% phone contacts BUT 54.5% WhatsApp contacts
• Profile: Active searchers who strongly prefer digital/text communication
• Behavior Pattern: Actively engaging via WhatsApp but avoiding phone calls
• Why they hesitate on RNPL:
  • May be comparing financing options
  • Could be waiting for right property
  • Possibly uncomfortable with phone-based verification process
• Recommendation:
  • DO invest in re-engagement via WhatsApp/digital channels
  • Simplify verification to reduce phone call requirements
  • Offer text-based application status updates
  • Consider in-app messaging for follow-ups

2. Focused Searchers (UNEMPLOYED)

• 60.4% make phone contacts with highest avg contacts (19.6)
• Profile: Serious intent, efficient searching despite employment barrier
• Behavior Pattern: Actively pursuing rentals, blocked by eligibility
• Recommendation:
  • High-value for re-engagement when employment status changes
  • Implement employment status tracking/notification system
  • Offer traditional rental options as alternative

3. Constrained but Engaged (DBR)

• 56.2% contact rate with strong engagement
• Profile: Financial constraints but genuine rental need
• Behavior Pattern: Want to rent, debt load prevents approval
• Recommendation:
  • Offer payment plan alternatives with different down payment structures
  • Consider longer term financing to reduce monthly burden
  • Re-engage after 6-12 months (debt may have decreased)

4. Credit-Limited Seekers (LOW_SIMAH_SCORE)

• Highest contact rate (65.8%) - most engaged segment
• Profile: Motivated searchers with credit history challenges
• Behavior Pattern: Actively pursuing rentals despite credit barriers
• Recommendation:
  • Top priority for partnership programs (credit building services)
  • These users are highly motivated and would convert with credit improvement
  • Consider secured/guaranteed rental products

User Classification Framework

Intent Classification Based on Phone Contacts

Intent Level	Criteria	Segment Example
High Intent	Made 10+ phone contacts	Core of UNEMPLOYED/DBR/LOW_SIMAH users
Medium Intent	Made 3-9 phone contacts	Mixed across all segments
Low Intent	Made 1-2 phone contacts	Edge of financially constrained
No Intent	Zero phone contacts	90% of HESITANT users

Recommended Re-engagement Priority (Revised)

Priority	Segment	Users	Channel	Rationale
1	LOW_SIMAH_SCORE	1,014	Phone/WhatsApp	Highest overall engagement (80%), addressable barrier
2	HESITANT	1,601	WhatsApp only	High WhatsApp engagement (55%), requires digital approach
3	DBR	1,268	Phone/WhatsApp	Multi-channel active (75%), financial may improve
4	UNEMPLOYED	1,735	Phone/WhatsApp	Multi-channel active (75%), status may change

Key change: HESITANT users moved UP in priority due to revealed WhatsApp engagement, but require WhatsApp-specific approach.

Recommendations

1. Channel-Optimized Engagement Strategy

HESITANT Users (Channel-Selective):

• Invest in WhatsApp-based re-engagement - 54.5% already engage via this channel
• Reduce phone call requirements in the RNPL process
• Offer text/chat-based verification alternatives
• Send WhatsApp reminders for incomplete applications
• Their “hesitation” may be process-related, not intent-related

Financially Constrained Users (UNEMPLOYED, DBR, LOW_SIMAH):

• High re-engagement potential - multi-channel active
• Phone follow-ups are effective (55-65% engage via phone)
• Implement status change tracking for employment/credit changes
• Offer alternative products (traditional rentals, roommate matching)

ACCEPTED Users (Phone-Preferred):

• Traditional phone-based communication works well
• May prefer formal/established communication channels
• Maintain current phone-first approach for qualified leads

2. Multi-Channel Pre-Application Qualification

Channel Activity	Intent Signal	Recommended Action
Phone + WhatsApp	Very High	Fast-track application
WhatsApp only	High (channel-selective)	Text-based follow-up
Phone only	High (traditional)	Phone-based follow-up
Neither	Low	Light-touch nurturing

3. Partnership Opportunities

• Credit Building: Partner with Simah improvement services for LOW_SIMAH_SCORE segment
• Employment Programs: Track/alert when UNEMPLOYED users gain employment
• Financial Planning: Offer DBR users debt consolidation resources
• WhatsApp Business: Implement automated WhatsApp flows for HESITANT segment nurturing

Application Persistence Analysis

Requests per User Distribution

Most users apply only once, but persistence varies significantly by segment.

# Requests	Users	%
1	7,911	85.3%
2	929	10.0%
3	264	2.8%
4+	167	1.8%

Average: 1.22 requests per user

Retry Rates by Segment

Segment	Users	Avg Requests	Single Try	Multiple Tries
DBR	1,220	1.45	73.0%	27.0%
LOW_SIMAH_SCORE	946	1.47	73.9%	26.1%
ACCEPTED	88	1.27	77.3%	22.7%
UNEMPLOYED	1,685	1.18	86.5%	13.5%
HESITANT	1,498	1.16	88.1%	11.9%

Key Persistence Insights

1. DBR and LOW_SIMAH_SCORE users are most persistent:

   • 27% and 26% make multiple attempts respectively
   • These users have financial constraints but strong intent to rent
   • Higher persistence indicates greater motivation despite barriers

2. HESITANT users rarely retry (11.9%):

   • Consistent with “window shopping” behavior identified earlier
   • Low retry rate validates they were never serious applicants

3. Most multi-request users receive the same rejection:

   • 123 users: UNEMPLOYED → UNEMPLOYED
   • 120 users: DBR → DBR
   • 87 users: LOW_SIMAH_SCORE → LOW_SIMAH_SCORE
   • Underlying issues don’t resolve quickly

Conversion to ACCEPTED

Metric	Count	Percentage
Total ACCEPTED users	90	-
Accepted on first try	68	75.6%
Needed multiple tries	22	24.4%

Rejection → Acceptance Conversions:

• 11 users converted from HESITANT to ACCEPTED (most common path)
• 4 users converted from OTHER_REJECTED to ACCEPTED

This suggests HESITANT users who do eventually convert had resolvable objections rather than fundamental disinterest.

SQL Queries Used

Pre-Application Phone Contacts by Segment

WITH user_contacts AS (
    SELECT user_id, count(*) as contacts
    FROM sadb_phone_get_logs FINAL
    WHERE _peerdb_is_deleted = 0 AND resource = 'listing'
      AND toDate(createdAt) >= '2025-04-01' AND toDate(createdAt) <= '2025-12-31'
      AND user_id IN (/* RNPL user IDs for segment */)
    GROUP BY user_id
)
SELECT
    'SEGMENT_NAME' as segment,
    TOTAL_USERS as total_users,
    count(*) as users_with_contacts,
    round(100.0 * count(*) / TOTAL_USERS, 1) as pct_with_contacts,
    sum(contacts) as total_contacts,
    round(avg(contacts), 1) as avg_contacts,
    quantile(0.5)(contacts) as median,
    quantile(0.9)(contacts) as p90
FROM user_contacts

Post-Application Retention (Jan 2026)

SELECT
    count(DISTINCT user_id) as users_with_jan_activity,
    count(*) as total_jan_contacts,
    round(avg(contacts), 1) as avg_contacts
FROM (
    SELECT user_id, count(*) as contacts
    FROM sadb_phone_get_logs FINAL
    WHERE _peerdb_is_deleted = 0 AND resource = 'listing'
      AND toDate(createdAt) >= '2026-01-01'
      AND toDate(createdAt) <= '2026-01-31'
      AND user_id IN (/* Rejected RNPL user IDs */)
    GROUP BY user_id
)

Data Sources

1. Ejari Requests CSV (ejari-aqar-requests.csv)

   • 7,153 applications (Aug-Dec 2025)
   • Contains: ID, Request Number, Status, Rejection Reason

2. RNPL Third Party Requests (sadb_rnpl_third_party_requests.sql)

   • 17,950 records
   • Maps request_id to Aqar user_id

3. Phone Get Logs (sadb_phone_get_logs)

   • 77M records (Jun 2021 - present)
   • High-intent engagement: user revealed landlord phone number
   • Used for phone contact rate analysis

4. WhatsApp Communication Flow (sadb_whatsapp_communication_flow)

   • 5.2M records (Dec 2024 - present)
   • Detailed WhatsApp conversation tracking
   • Used for WhatsApp engagement analysis

5. User Listing Views (sadb_user_listing_views)

   • 1.26B records (Feb 2022 - present)
   • Browsing behavior data
   • Used for view activity sampling

Limitations

1. Channel overlap not deduplicated: Users engaging via both phone AND WhatsApp are counted in both metrics. The “Either (est)” column approximates combined reach but may overcount.

2. WhatsApp data recency: WhatsApp data starts Dec 2024, so earlier RNPL applicants may have incomplete WhatsApp history.

3. Pre-application window approximation: The 30-day window before application uses cohort-level date filtering rather than per-user exact dates due to query performance constraints.

4. Post-application window: Jan 2026 retention data is limited to ~1 month post-rejection for December applicants.

---

Analysis conducted by Claude on February 2, 2026 Updated with complete historical data coverage (7,812 users vs previous 2,174)