1 Summary
Marketing data reveals clear product improvement opportunities.
APP CVR
1.22%
View Home → Complete Booking
WEB CVR
0.18%
View Tour Detail → Complete Booking
Key Insight
The 7x CVR gap is not a marketing problem — it's a product UX problem.
Users who understand TOMOGO! through the web before installing convert at 5-7x higher rates in-app. Whether they arrive "educated" or not determines everything.
2 APP: Onboarding is the Biggest Bottleneck
Data
| Metric |
Direct Install |
WEB Conv → Install |
Gap |
| AppContentView / Install |
96-127% |
568-760% |
5-7x |
| AppSearch / Install |
56-60% |
239-320% |
4-5x |
Insight
Users who land directly in the app from install ads don't understand what TOMOGO! offers. They browse minimally (1 tour view per install) and leave. Users who first visited the website understand the value proposition, install intentionally, and
browse actively (6-8 tour views per install).
User Flow Comparison
Direct Install (current majority):
Ad Click
→
Install
→
"What is this?"
→
Drop-off
WEB Conv → Install (ideal flow):
Ad Click
→
Browse Web
→
Understand Value
→
Install
→
Active Usage
Recommendation
- App onboarding flow should replicate what the website does: introduce TOMOGO!'s unique value (local guides, hidden spots, customizable tours)
- First-open experience should showcase: top tours, tour leader profiles, reviews/testimonials
- 46-51% of users book within 0-1 day of account creation — the first session is everything
- Push notification within 24h for new installs with personalized tour recommendation
3 APP: Fix the Tour Leader Selection Loop
Data
r = -0.82
Meta CartAdd vs Bookings Correlation
Negative
More confirmation views = Fewer bookings
Problem
More users reaching the confirmation screen correlates with
fewer actual bookings (r = -0.82).
Root cause: Users reach confirmation → see Tour Leader selection → go to TL Detail Page → come back → try another TL → repeat loop (decision paralysis).
Current Flow (Problem)
Tour Detail
→
Book Now
→
Confirmation
TL selection
→
TL Compare Loop
→
Drop-off
Improved Flow (Proposed)
Tour Detail
TL profiles shown
→
TL Decided
→
Book Now
→
Confirmation
TL info only
→
Booking Complete
Recommendation (aligns with dev team's existing plan)
- Move Tour Leader selection/preview BEFORE the checkout flow, not during it
- Show TL profiles on the tour detail page so users decide before clicking "Book Now"
- On the confirmation screen, show only the assigned TL with a brief bio — not a selection interface
4 WEB: Fix 91% Funnel Drop-off
Web Funnel Data (Amplitude 4/10-4/26)
541 UU
View Tour Detail Page
48 UU
Complete Checkout Date Time
91.1% drop
1 UU
Complete Booking
CVR 0.18%
APP vs WEB Funnel Comparison
APP
1.22%
1,807 UU → 22 Bookings
WEB
0.18%
541 UU → 1 Booking
Session Replay Behavioral Patterns
| Behavior |
% |
Implication |
| Navigate to another tour detail (comparison fatigue) |
58.6% |
No related tour suggestions |
| External traffic (no way to browse back) |
37% |
Missing browse navigation |
| Auth Dialog completion rate |
10% |
Authentication is a major barrier |
Recommendations (Priority Order)
- Add "Related Tours" section (3-5 same-area tours) to tour detail pages — #1 priority from dev team analysis
- Implement guest checkout (confirmed in 4/27 meeting) — Auth Dialog's 10% completion rate is critical
- Add "Recently Viewed" section for returning visitors
- Improve calendar loading speed (functional issue)
- Suppress app download banner after first dismissal per session
5 WEB: Maximize WEB-to-Install Path
Data
685%
WEB Conv → Install AppCV Rate
96%
Direct Install AppCV Rate
4.6/day
Avg Bookings on WEB Purchase Days
2.3/day
Avg Bookings on Normal Days
Insight
The website is functioning as an "education layer."
Users who browse the site before installing are pre-qualified and convert at dramatically higher rates. The site's role is not just direct conversion — it's
creating high-quality app users.
Recommendation
- Ensure "Download the App" CTA is prominent on all tour detail pages (but suppress after dismissal)
- Consider a "Save this tour" feature that bridges web to app (deeplink to specific tour in app)
- Track the web-to-install-to-booking journey as a key funnel metric
- Optimize web tour detail pages for "understanding value" rather than "completing checkout" — the real conversion happens in the app
6 Content: Browsing Experience Design Guidelines
Content Performance Data
| Tour / Category |
Characteristic |
Product Implication |
| Secret Tokyo |
#1 in both views and bookings |
Position as anchor content prominently |
| Omakase (customizable) |
Scales best in high-demand periods |
Expand personalization features |
| Asakusa |
Consistent performer across all months |
Evergreen content — always visible |
| Bar Hopping |
High male engagement, reactivation of existing users |
Hook for retention campaigns |
Content Priorities for WEB and APP
- Feature Secret Tokyo / Omakase prominently at the top
- Increase Tour Leader profile visibility — TL-fronted content converts better (e.g., Yanaka Ad 74% female, Rin 2 Ad 63.8% female)
- Area-based browsing (related tours in same area) — matches user behavior (58.6% navigate to another tour)
- "Local insider" content theme — SNS analysis shows it drives bookings 7-14 days later
6b Coupon UX Improvement — 75.7% of First Bookings Use No Coupon
75.7% of First Bookings (305/403) are made without any coupon. The ebr20 coupon is delivered via in-app chat (Admin→Guest message), but users have no reason to check chat immediately after install, resulting in 90%+ oversight rate.
Coupon Distribution Channels
| Channel | Method | Usage Rate | Issue |
|---|
| ebr20 | Auto-sent via in-app chat (Admin message) at Install | 9.2% of First BK | No motivation to open chat. 90%+ missed |
| WEL20 | WELCOME email after CA | 2.5% of First BK | Email volume -82% after CA flow change |
Improvement Recommendations
| # | Area | Current | Recommended |
|---|
| 1 | Coupon visibility | Hidden in chat box only | Add "Unused Coupon" section to profile/my page. Home screen badge notification |
| 2 | Auto-apply in booking flow | Manual coupon code entry required | Show "Apply 20% OFF" one-tap button on checkout if unused coupon exists |
| 3 | LP/Website coupon promotion | No coupon info on website | "First tour 20% OFF in the app" banner on tour detail pages. Drives Install + coupon awareness |
| 4 | Push notification reminder | MoEngage does not promote coupons | Scale "First coupon reminder" Push (CVR 14.3%) to all CA users. Remind at Day 3/7 post-install |
| 5 | Cart abandon Push expansion | 4 deliveries/month. CVR 25.0% | Relax trigger conditions to scale volume. Highest ROI Push campaign |
MoEngage Push shows 2.5x CVR (9.1% vs 3.6% baseline) but only 5 bookings/month. Behavioral trigger Pushes (cart abandon CVR 25%, first coupon CVR 14.3%) are the highest-ROI improvement opportunity. MoEngage is used exclusively for re-engagement, not coupon distribution, but adding coupon reminders to Push notifications has significant potential.
7 Measurement Improvement Proposals
Current Gaps
| Issue |
Impact |
Status |
| UTM tracking broken since 1/16 |
AppsFlyerOneLink × Amplitude mapping incomplete |
In progress (large project) |
| Wix LP → main site traffic not tracked |
Incomplete web traffic picture |
Not started |
| Amplitude has no gender data for app users |
"Women discover → men book" hypothesis unverifiable in-app |
Not started |
| Web funnel data only available since 4/10 |
Need 30+ days for reliable analysis |
Accumulating |
Improvement Priority
- Fix UTM mapping (Miyanishi team — large project, in progress)
- Add gender property to Amplitude user profiles
- Implement cross-platform journey tracking (Web session → App install → App booking)
- Set up weekly automated funnel report (Web + App side by side)