TOMOGO! Weekly Performance Analysis

1. Executive Summary

The strong performance on 4/16-17 was driven by a confluence of temporary factors (SNS post surge + 1 user's bulk booking of 6 tours + initial CVs from new ads) and does not represent structural growth.
On a unique-user basis, 4/18-19 (2-3 users/day) is on par with 4/17 (2 users). The apparent sharp decline is misleading.
Organic post ads exhaust their CVs within D+0 to D+1, with zero CVs from D+2 onward. Four out of six historical spikes reverted within two weeks.
The true value of boosted posts lies not in direct CVs but in awareness that drives organic traffic the following week (r = 0.35-0.40).

2. Daily Performance Trend (4/13 - 4/19)

Date	Bookings	First	Repeat	UU	Pax	Revenue	Ad Cost	CVR	ROAS	SNS Views
4/13 (Mon)	3	3	0	3	7	¥82,610	¥98,293	0.65%	—	0
4/14 (Tue)	0	0	0	0	0	¥0	¥89,702	0%	0%	0
4/15 (Wed)	2	2	0	2	3	¥32,697	¥92,206	0.84%	35.5%	0
4/16 (Thu) Peak	5	4	1	4	14	¥168,080	¥155,688	2.65%	94.1%	34,429
4/17 (Fri) Peak	7	1	6	2	14	¥194,920	¥159,496	2.71%	97.2%	11,625
4/18 (Sat)	2	2	0	2	3	¥42,680	¥164,293	0.66%	26.0%	0
4/19 (Sun)	3	3	0	3	4	¥65,956	¥162,935	0.86%	40.5%	0

3. 4/17 Booking Breakdown — Repeat-Heavy Reality

Detailed breakdown of the 7 bookings on 4/17:

#	Type	Channel	Tour	Pax	Value	UserID
1	Repeat	App	Asakusa: Tokyo's Timeless Charm	2	¥21,560	****7134
2	First	App	Omakase Osaka	2	¥39,820	****2055
3	Repeat	App	Flavors of Osaka	2	¥25,300	****2055
4	Repeat	App	Omakase Kyoto	2	¥39,820	****2055
5	Repeat	App	Kyoto Walking Tour	2	¥21,560	****2055
6	Repeat	App	Shibuya's Hidden Gems	2	¥21,560	****2055
7	Repeat	App	Fresh Flavors of Tsukiji	2	¥25,300	****2055

UserID ****2055 alone booked 6 tours (86% of the day's total). This is a typical pattern of itinerary bulk-booking.
This has occurred repeatedly in the past (Nov 2025: 3 users booked 11 tours total; Jan 2026: 1 user booked 7 tours).

Unique User Comparison

Date	Bookings	Unique Bookers	New Users
4/16	5	4	4
4/17	7	2	1
4/18	2	2	2
4/19	3	3	3

On a unique-user basis, 4/17 had the fewest bookers. 4/18-19 actually acquired more new users. Judging by booking count alone is misleading.

4. New Organic Post Ad Effectiveness Analysis

19 new ads (tour-name SNS post format) were launched on 4/16.

4a. New vs. Existing Ads — Daily Comparison

Date	New IMP	New CV	Existing IMP	Existing CV
4/13	0	0	31,332	0
4/14	0	0	28,771	0
4/15	0	0	28,036	0
4/16	24,245	2	28,141	1
4/17	25,545	1	23,502	0
4/18	36,270	0	21,495	0
4/19	41,512	0	27,001	0

New ad impressions keep growing, but CVs were exhausted by D+1 — zero from D+2 onward.
Existing ads (Conversion_Main, etc.) show gradual IMP decline from 4/16 onward — potential budget cannibalization.

4b. Full-Period Data: Organic Post Ad CV Decay Pattern

Aggregate of all 64 ads (ad_raw basis):

Days Since Launch	Active Ads	Total IMP	Total CV	CVR
D+0	64	111K	2	0.24%
D+1	62	253K	1	0.04%
D+2	63	255K	0	0.00%
D+3	55	282K	0	0.00%
D+4 to D+14	12-38	68-179K	0	0.00%

Only 5 out of 64 ads (7.8%) ever produced a CV. The vast majority of organic post ads never convert.

5. Weekly Trend — Post-Spike Revert Pattern

Performance trajectory after 6 spike events where organic post ad IMPs surged:

Spike Week	Spike Bk/day	+1w	+2w	+3w	Result
2025-W42 (Oct)	2.4	1.7 (-29%)	1.6 (-35%)	2.1	REVERT
2025-W46 (Nov)	2.7	3.9 (+42%)	6.1 (+126%)	2.9	GROW then REVERT
2026-W04 (Jan)	1.9	5.3 (+185%)	2.3 (-57%)	2.3	+1w only then REVERT
2026-W12 (Mar)	3.3	2.7 (-17%)	1.6 (-52%)	1.6	REVERT
2026-W16 (Current)	3.1	?	?	?	REVERT predicted based on pattern

4 out of 6 spikes reverted within 2 weeks. The only sustained case (Nov) coincided with peak inbound tourism season.

Correlation Data

Correlation	r-value	Interpretation
Org IMP → Same-week Bookings/day	0.279	Weak positive correlation
Org IMP → Next-week Bookings/day	0.350	Moderate positive
Org IMP → Next-week UU	0.396	Moderate positive

The boost's value lies not in direct CVs but in awareness that lifts organic traffic the following week.

6. Recommendations

1. Redefine organic post ads as "awareness boosters"

These are not CV-generating ads — treat them as awareness campaigns and reallocate budget accordingly. Ensure they do not cannibalize the Conversion_Main budget.

2. Add unique-user-based metrics to CPO/ROAS evaluation

Booking-count-based CPO fluctuates heavily due to repeat bulk bookings. Evaluate true acquisition efficiency using Unique Bookers ÷ Ad Cost.

3. Judge new ads within 3 days — redirect budget to existing CV campaigns if CV is zero

The D+2 zero-CV pattern is now confirmed. Leaving underperforming ads active reduces Conversion_Main delivery volume.

4. Shift 15-20% of weekend ad spend to weekdays

Saturday has the lowest bookings of any day (avg. 1.4/day). Same spend yields 1/3 the CVR. Shifting to Thu-Fri is more efficient.

5. Expand Apple Search Ads (ASA) budget

CPI (Cost per Install) of ¥4,161 is efficient. However, ASA CVs are app installs, not booking conversions — they cannot be directly compared to other channels' CPO (cost per booking). Measure Install-to-Booking conversion rate before scaling budget.