Execution Roadmap

Global Onboarding
UI/UX Improvements

Prioritized by funnel drop-off percentage. From 34.6% to 45% conversion target. Focused on UI-actionable steps only.

📊
81,384 users
Viewed Onboarding
✅
28,156 approved
34.6% conversion
🎯
Target: 45%
+10.4pp to close gap
📅
March 19, 2026
Product & Ops Team

Context

The Global onboarding funnel converts 34.6% of users from "Viewed Onboarding" to "Approved" — 10.4 percentage points below the 45% target. This roadmap organizes UI/UX improvements by drop-off severity, focusing on steps where design changes can directly reduce abandonment.

Backend/processing steps (Onfido Check, Pending Review, Bank Approved) are excluded — this document is for the person implementing UI changes.

Priority Overview

Ordered by drop-off percentage — highest first

P1
Account Created + Address Input
22,426 users lost
27.6%
View
P2
ID Number Entry
7,506 users lost
12.8%
View
P3
Bank Requested
6,495 users lost
12.8%
View
P4
Start Onfido (Photo Guidance)
1,621 users lost
3.7%
View
P5
Document Type Selection
558 users lost
0.9%
View
P1

Account Created + Address Input

CRITICAL — Largest funnel leak
27.6%
Drop-off
22,426
Users Lost

What happens here

Users create a basic account and explore features (virtual card, balance, "My Panas") before completing KYC. The user sees 9 screens of features in Account Created before reaching Address Input. When asked for their physical address, 27.6% abandon. The gap between "exploring cool features" and "now give me your personal info" creates friction.

Drop-off by Country

Country Users Drop Rate Status
🇧🇴 Bolivia 4,882 20.5% Best
🇭🇳 Honduras 5,998 24.4% Acceptable
🇩🇴 Rep. Dominicana 36,125 24.6% #1 Market
🇨🇴 Colombia 11,656 26.0% Near average
🇲🇽 Mexico 10,020 27.5% Average
🇬🇹 Guatemala 8,631 34.6% Critical
🌎 Otros (~88 paises) 3,897 56.5% Highest Drop

Platform Gap

Android
31.3%
abandonment
iOS
20.8%
abandonment
⚠️
10.5pp gap — suggests UX problems on Android, especially on low-end devices common in LATAM (Samsung A-series, Motorola G-series, Xiaomi Redmi).

Recommended UI Changes

A — High Impact
Address Form Localization
  • Guatemala is missing the "Zona" field — essential for GT addresses. Likely explains the 34.6% drop.
  • Implement country-specific fields: Zona (GT), Colonia (MX), Barrio (HN), etc.
  • Auto-detect user's country and adapt the form accordingly.
B — High Impact
Reduce Exploration-to-KYC Gap
  • Show a clearer "complete your verification" path instead of 9 screens of features.
  • Current flow sets wrong expectations — users explore features they can't use yet.
  • Recommendations already reviewed with Nahomi for implementation.
C — Medium Impact
Android-Specific Optimizations
  • Review keyboard/autocomplete behavior on low-end Android devices.
  • Test on Samsung A-series, Motorola G-series, Xiaomi Redmi.
  • Google Places may need manual-entry-first on slower devices.
D — Medium Impact
Address Autocomplete
  • Current: Google Places with "Can't find your address?" manual fallback.
  • Add country-specific formatting and support for informal LATAM addresses.
Expected Impact

If drop-off reduces to Bolivia's level (20.5%), that's ~5,800 additional users entering the funnel.

P2

ID Number Entry

MODERATE but WORSENING — Urgent for Central America
12.8%
Drop-off
7,506
Users Lost

What happens here

Users enter their passport or national ID number and expiration date. The form is the same generic layout for all countries — no format guidance, no document image reference, no adaptation to local ID formats.

Worsening Trend — Central America Critical

2025 overall: 10.5% abandonment (6,202 users lost). Rate increased from 9.0% (Q3) to 11.0% (Q4).

Country Drop Rate Trend
🇲🇽 Mexico Improved -7.4pp (fixes working)
🇭🇳 Honduras 16.1% +12.5pp worse
🇬🇹 Guatemala 13.3% +8.0pp worse
🇨🇴 Colombia 14.3% +6.8pp worse
✅
Mexico's -7.4pp improvement proves that targeted fixes work. The same approach needs to be applied urgently to Central American markets.

Passport vs National ID

Passport
14.2%
abandonment (9.9% of users)
National ID
7.2%
abandonment (90.1% of users)

Passport has almost 2x more abandonment than national ID.

Current Problems

# Problem Impact
1 Generic form for all countries — no adaptation to local document formats High
2 No format example — user doesn't know what to type (e.g., DR cedula is XXX-XXXXXXX-X) High
3 No document image reference — can't identify where the number is Medium
4 Security message at the bottom — should appear before the sensitive input Medium
5 Expiration date asked for all countries — some IDs don't expire Medium

Recommended UI Changes

A — Highest Impact
Country-Adaptive Form
  • Auto-detect country from previous steps
  • Show correct ID format with example (e.g., "XXX-XXXXXXX-X" for DR cedula)
  • Show reference image of the document highlighting where the number is
  • Hide expiration date for countries where the document doesn't expire
B — Medium Impact
Real-time Validation
  • Validate format as user types (length, characters, check digit)
  • Show green checkmark when format is valid
  • Reduces anxiety about "did I type it right?"
C — Medium Impact
Trust-First Layout
  • Move encryption/security message ABOVE the input field
  • Add "Why do we need this?" expandable explanation
  • Show lock icon next to the input field
D — Passport-Specific
Passport Improvements
  • Add passport image reference showing where number is
  • Include real example of passport number format
  • Single date picker instead of DD/MM/YYYY split fields
Expected Impact

If Central America returns to Q3 levels + passport drop cut by half: 12.8% → ~8-9%, recovering ~2,000-2,500 users.

P3

Bank Requested

MODERATE — Single field causing most friction
12.8%
Drop-off
6,495
Users Lost

What happens here

Users fill out compliance data for Bridge (Stripe): Account Purpose, Employment Status, Occupation, and Source of Funds. Then accept Bridge Terms & Conditions and Privacy Consent. Total: 3 screens.

The Core Problem: 130 Occupations in a Dropdown

The "Most Recent Occupation" field has 130 options in a flat dropdown with no search, no categories, and very technical English titles (e.g., "Adhesive Bonding Machine Operator").

Account Purpose
10
OK — Manageable
Employment Status
6
OK — Easy
Source of Funds
11
OK — Moderate
Occupation
130!
CRITICAL
⚠️
Critical countries: Guatemala 19.1% and Honduras 16.1% drop rate at this step.

Occupation Field Solutions

1. Simplified Categories
  • Reduce to 20-30 generic categories
  • "Technology", "Healthcare", "Retail", "Construction"...
  • Map internally to Bridge's 130 options
  • Users see simple, recognizable labels
Impact: -5 to -7% abandonment
2. Search + Categories
  • Search input with autocomplete
  • Group options by category
  • Show top 10 most common first
Impact: -4 to -5% abandonment
3. Smart Suggestions
  • "Employed" → show common jobs
  • "Self Employed" → freelance options
  • "Student" → auto-select
Impact: -3 to -4% abandonment

Additional Improvements

B — Trust
Bridge Terms Co-branding
  • Current: only Bridge/Stripe logo — users may not recognize it
  • Add Pana logo alongside with "Powered by Bridge (Stripe)" messaging
  • Builds trust and explains the partner relationship
C — Trust
Trust Elements
  • Security badge (256-bit encryption) on all input screens
  • Progress indicator ("Step 2 of 3")
  • "Why do we need this?" tooltips
  • Partner logos in footer
D — UX
Form UX
  • Visual card options with icons for Account Purpose
  • Chips/pills for Employment Status (6 options)
  • Progress bar across the 3 screens

Expected Impact per Improvement

Improvement Est. Reduction Justification
Occupation search + autocomplete -3 to -4% Eliminates scrolling 130 options
Grouped categories -1 to -2% More intuitive navigation
Translations + simplified labels -2% Guatemala/Honduras improve
Visual options with icons -1% Reduces cognitive load
TOTAL ESTIMATED -7 to -9% 12.8% → ~5%
P4

Start Onfido (Photo Guidance)

GOOD — But quality issues cause downstream rejections
3.7%
Drop-off
1,621
Users Lost

What happens here

Users go through the Onfido SDK: document photo instructions, front/back capture, selfie instructions, selfie capture, and review. 6+ screens handled by Onfido. Despite complexity, only 3.7% drop.

Why it matters beyond 3.7%

Poor photo quality here causes rejections at Onfido Check (14.6% drop) and Bank Approved (18.5% drop). The #1 cause of failed verification is photo quality (35%), especially on Android low-end devices.

Common Failure Causes (at verification)

35%
Photo quality
(blurry, bad lighting)
25%
Document
verification failed
18%
Face mismatch
(selfie vs ID)
12%
Expired
documents
10%
Other
causes

Recommended UI Changes

A — High Impact
Pre-capture Quality Validation
  • Validate lighting conditions before "Take photo"
  • Real-time quality indicator (red/yellow/green) for lighting
  • Block capture if image is too blurry
B — Medium Impact
Better Positioning Guide
  • Overlay frame with clear corners for alignment
  • Real-time: "Move closer", "Hold steady", "Better lighting"
  • Guidance for cedula size vs passport size
C — Medium Impact
Selfie Improvements
  • Face oval guide with alignment feedback
  • "Face a window", "Avoid backlight" guidance
  • Ensure selfie matches document (no glasses changes)
D — Low Impact
Retry Without Restart
  • Allow retaking just one photo without restarting entire flow
  • Partially handled by Onfido SDK
  • Reinforce with messaging
Expected Impact

Direct: -0.5 to -1% on this step. But the indirect impact on downstream rejections (Onfido Check 14.6% + Bank Approved 18.5%) could be significant. Better photos = fewer rejections = more approved users.

P5

Document Type Selection

HEALTHY step — Strategic change needed
0.9%
Drop-off
558
Users Lost

What happens here

Users select their nationality and choose between Passport or National ID. Single screen, simple binary choice. 0.9% is excellent.

Why it's in this roadmap

The step itself is fine, but the choice users make here has major downstream consequences:

Passport Users
14.2%
abandonment at ID Number step
National ID Users
7.2%
abandonment at ID Number step

Passport has 2x the abandonment of national ID. Only 9.9% of users choose passport, but they abandon at much higher rates. Many passport users from DR get rejected because passport verification is stricter.

Recommended UI Changes

A — Quick Win
Pre-select National ID
  • 90.1% of users choose National ID anyway
  • Pre-selecting saves a tap and guides toward higher success path
  • Passport remains fully available
B — Subtle Nudge
Success Rate Context
  • Next to National ID: "Recommended — Fastest verification"
  • Not about hiding passport, but helping users make the best choice
  • Gentle nudge toward cedula in DR market
Expected Impact

Minimal on this step's 0.9%, but downstream impact could be significant. If 50% of passport users switch to national ID, that's ~2,900 users with a better chance of completing verification.

Total Potential Recovery

34.6%
Current Conversion
→
42-45%
Target Conversion
Priority Step Current Target Users Recovered
P1 Account Created + Address 27.6% ~21% ~5,400
P2 ID Number Entry 12.8% ~8.5% ~2,200
P3 Bank Requested 12.8% ~5% ~3,500
P4 Start Onfido 3.7% ~3% ~300 (+ downstream)
P5 Document Type 0.9% ~0.9% downstream impact

Key Markets to Watch

Prioritized by user volume and conversion opportunity

🇩🇴
Rep. Dominicana
#1 Market
36,125 (45%)
Core market — every 1% improvement has massive impact
🇨🇴
Colombia
11,656 (15%)
Second market — monitor ID Number trends
🇲🇽
Mexico
10,020 (13%)
Improving with fixes — replicate success
🇬🇹
Guatemala
Critical
8,631 (11%)
Worst performer in Address + Bank Requested
🇭🇳
Honduras
Critical
5,998 (8%)
Worsening in ID Number — urgent attention
🇧🇴
Bolivia
Best
4,882 (6%)
Best performer — study what works here

Demographics Context

Who are Global users — design decisions should reflect this

40.5%
Age 25-34
26,969 users
25.3%
Age 18-24
16,860 users
22.8%
Age 35-44
15,192 users
11.5%
Age 45+
7,622 users
💡
63% are under 35 — digital-native, mobile-first, low patience for friction. Low-end Android devices are the primary platform in these LATAM markets. UI must be modern, fast, and mobile-optimized.

Resources

Links to dashboards, designs, and documentation

📊
Global Dashboard
✨
UX Optimization Page
🎨
Figma Design
📝
Notion Documentation
📈
Amplitude Funnel
🏦
Bridge API Docs
Pana Global — Onboarding UI/UX Execution Roadmap — March 2026