Idle Cat Gunner Ads: Quiet for Months, Then a September Hit
A cat shooter that sat quiet for months turned September into a ramp: 2,930 ad plans in 31 days, 1,595 new creatives in two months, three markets bought alone.
Daniel Brooks
Creative Strategy Analyst
Published
Updated
A cat shooter that sat quiet for months turned September into a ramp: 1,595 new creatives pushed into rotation across four parallel networks, with three markets bought alone.
The data scope
Every figure in this breakdown comes from CreativeScope’s ad-intelligence dataset for Idle Cat Gunner: Shooter RPG (publisher listing: TREEPLLA), measured over the 31-day window 2026-09-09 to 2026-10-09. A plan is one ad delivery setup; a creative is the underlying video, image or playable asset a plan points at. “Exclusive” means a plan whose target-country list contains exactly one market — a targeting decision, not a performance measurement.
Two scope notes before the numbers. First, the store record carries no usable launch date, and the 1M+ downloads label puts this well past day one; we read the buy as steady-state live acquisition rather than a launch. One chart reaches further back than the window: a 61-day series from 2026-08-10 to 2026-10-09 tracks the daily creative pace and uses the same definitions as the rest of the article. Second, CreativeScope also carries a second record for the same title under a related publisher entry — we report it separately in its own section and never add the two books together.
The game in one paragraph
Idle Cat Gunner: Shooter RPG is a side-view idle shooter from TREEPLLA, a Seoul-based studio founded in May 2021 that operates under Neptune, a company with its own KRAFTON investment history. A squad of round, dead-eyed cats marches through biome maps and fires on its own; the player’s real job is the mergeable weapon inventory, which doubles as the progression spine. The store listing carries a 1M+ downloads label. Everything about the buy — 2,930 plans in 31 days across 13 networks — says an operator past product-market fit, spending to keep the funnel full.
A two-month creative surge
Every figure in the structural sections that follow describes a 31-day window. Widen the lens to the 61 days from August 10 to October 9 and a pattern appears that a single month hides: the creative pipeline was not steady, it was climbing. CreativeScope logged 1,595 new creatives across those two months, and the daily rate roughly doubled from start to finish.
New creatives first recorded by CreativeScope on each calendar day, August 10 to October 9, 2026. The orange line is a trailing seven-day average of the daily count. The dashed rule marks September 9, where the 31-day window used in the rest of this article begins. A creative counts on the day it is first seen; this series is context and is not added to the window figures elsewhere in the article.
August averaged 18.5 new creatives a day. September averaged 26.8. The first nine days of October averaged 42.6, and the trailing seven-day average peaked at roughly 50 in the final week. Counted in weekly blocks, the register steps from 114 creatives in the week of August 10 to 347 in the week of September 28 — closer to a threefold climb than a drift.
The climb is lumpy rather than smooth. The three heaviest days of the entire series — September 30 (71 creatives), October 3 (76, the two-month peak) and October 7 (66) — all fall inside the final ten days, and no day in August comes close. Read beside the structural sections below, the picture is a buy that was still accelerating at the end of the measured period rather than tapering toward it.
Three markets buy alone; everyone else rides one flat track
The market table splits cleanly into two shapes. Korea, the US and Japan are bought separately: Korea has 835 plans targeting it, 752 of them exclusive to Korea (90%); the US has 673 plans, 381 exclusive (57%); Japan has 570 plans, 380 exclusive (67%). Add them up and 1,513 plans — more than half the entire book — were pointed at exactly one of those three countries and nowhere else.
Then there is the flat track. Canada (593 plans), France (593), Spain (579), Germany (568), Thailand (541), Australia (511) and Poland (503) all cluster in a 90-plan band, and their exclusive counts collapse to single or low double digits: 18, 13, 12, 7, 9, 11 and 1 respectively. The same creatives saturate all of them. No other market in the top ten gets its own budget line.
Bar length = ad plans targeting the market on a fixed axis (full bar = 900 plans); orange = plans exclusive to that market, teal = plans shared with other markets. Values show coverage · exclusive count.
The copy in the dataset makes the split visible, not just the counts. The shared Meta package runs English copy (“Tiny paws, MIGHTY guns! Cutest idle cat shooter RPG…”). Sitting next to it is a Korean lane: at least four distinct Meta images carrying Korean ad copy (roughly “Destroy everything together with your cat”), a Korean-subtitled video with a local disclosure line baked into the corner, and even an AppLovin playable preview card with a Korean headline. Japan gets its own video with a Japanese title that transliterates as “Nyanko Gunslinger” and copy reading roughly “crush it with overwhelming firepower.” The US lane shows up as a targeting fact — 381 plans bought for the US alone — rather than as visibly localized copy in this sample.
Four lanes running side by side
Most records we read lean on one or two networks and treat the rest as leftovers. This one runs four meaningful lanes in parallel: the Meta four-surface package (Audience Network, Facebook, Instagram and Threads each cover 42.39% of plans, Messenger 32.46%), Google’s AdMob at 21.81%, AppLovin at 16.18%, and a native-ad tier of Vungle (8.57%), ironSource (4.06%) and Unity (3.96%). Thirteen networks appear in the window overall. The shares are coverage rates with overlap — one plan can hit several surfaces — so they do not sum to 100%.
Bar length equals the printed share on a true 0–100% axis (a 42.39% bar fills 42.39% of the track). Coverage is multi-attribution: one plan can appear on several networks, so shares overlap and do not sum to 100%. Eight further networks each below 4% are omitted.
Each lane also gets its own asset class. Meta carries the mainstream video-and-image package in five sizes. AdMob takes the landscape odds and ends. AppLovin is where the playables and the highest-exposure standalone videos live. The native tier (Vungle, ironSource, Unity) recirculates the same vertical videos into interstitial and rewarded slots. That is four buying systems maintained at once — for a studio of roughly fifty people, per LinkedIn.
What the 1,331 creatives are made of
Counting by network attribution (a creative on three networks counts three times), the window holds 1,399 creative records carrying 2,930 plans. Videos dominate: 1,125 records (80.4%) covering 2,399 plans. Images are a thin slice — 100 records, 298 plans — almost all of them in the Korean Meta lane. Playables are the quiet third pillar: 174 records for 233 plans, effectively all on AppLovin.
- Format: 1,125 video / 100 image / 174 playable records — 80.4% / 7.1% / 12.4% of attributed creatives.
- Orientation: 1,108 of the 1,125 videos (98.5%) are vertical; only 15 landscape videos exist, and they run almost exclusively on AdMob (14 records) and YouTube (11).
- Where the volume sits: by attribution, AppLovin hosts 427 creative records (16.3%), each Meta surface 291 (11.1%), Messenger 290, AdMob 233 (8.9%), Vungle 182 (7.0%), then Unity (97), ironSource (92), YouTube (53), Mintegral (42) and TikTok (35).
- The top of the heat list is AppLovin-shaped: of the twelve highest-heat creatives in the window, nine are AppLovin assets — webm videos and playables whose estimated exposures sit between 1.12M and 1.48M each.
The second listing next door
CreativeScope carries a second record for Idle Cat Gunner under a related publisher entry (Neptune Company). It is a book of similar shape but roughly a third of the size: 1,038 creatives, 1,243 plans, 38.8M lifetime estimated exposure. We report it here as a single sentence and keep it out of every calculation above — the two listings are never added together, and every number in this article refers to the TREEPLLA record unless explicitly stated. Readers reconciling our figures against their own dashboards should decide which book they are looking at before comparing.
Creative spotlight
The mainstream package. This is the creative the whole Meta system is built around, and it is a three-act structure. It opens on a deadpan two-second cartoon — a tree with an unimpressed face — that has nothing to do with the game and everything to do with stopping the scroll. It then cuts to the merge inventory: a grid of duplicate pistols the player fuses into stronger guns, the exact loop the ad wants you to imagine doing. The last act is the payoff fantasy: a lava biome where the Lv.99 crowned cat mows down waves while damage numbers spike from 246.17A to 3.39B. A title card with a “Download now!” button closes it. The same family ships in at least 15 edits across Audience Network, Facebook, Instagram, Messenger and Threads.
The AppLovin boss-fight. The single biggest asset of the window leads with its end card: the logo and a “Download now!” button hold the first seconds before any gameplay appears. It then moves to a cherry-blossom canyon where a Lv.10 cat in a red jacket stands alone, and resolves into the ad’s one argument — a Lv.10-versus-Lv.1 confrontation staged like a boss fight, with the smaller cat visibly outgunned. No captions, no voiceover; the contrast does the selling. This is the lane where TREEPLLA spends on quality: 720p, a single-edit family, live the entire window.
The TikTok story cut. The same game, told as a sixty-second narrative with white caption overlays in the platform’s confessional style. It starts at the very beginning — a coin counter at 98, a cat shooting trees for cash — then hits a shop scene captioned “take this B-rank rifle,” a crate-smashing beat captioned “ONE MORE,” and closes on a full-screen blue beam under the words “999 BILLION POWER.” It is the only creative in the sample that argues progression over time rather than in a single image, and the only one wearing TikTok-native clothes.
The Korea-only cut. A fifteen-second edit with Korean caption overlays on every scene — desert, forest, lava — and a small Korean disclosure line pinned to the bottom-right corner, the kind local regulation expects on chance-based item messaging. The record ships under a dynamic template variable rather than fixed copy, runs in a family of eight edits, and never appears in the shared English package. This is what a 752-plan exclusive buy looks like at the asset level: not a translated subtitle slapped on the global video, but a separately produced cut for one market.
The playable tier. The AppLovin lane is not only video. Its playable program runs 174 creative records for 233 plans across the window, and the largest of them — previewed here with a Korean headline over a three-biome split — carries an estimated 1.42M exposures, more than any Meta video in the sample. A sibling record with the identical preview has run since mid-July and adds another 1.26M. For a studio running four lanes at once, the interactive lane is not an experiment; it is a standing register with its own renewal cadence.
A pulse, not a firehose
The window has no single launch spike. New volume arrives in pulses, and the two biggest days tie at 154 plans: September 30 (71 fresh creatives) and September 17 (45). October 1 adds 133 plans on 53 creatives, and the October 7 pulse (110 plans, 66 creatives) shows the buy was still live on the last full day of the window. One date breaks the pattern in an interesting way: October 3 shipped 76 creatives for only 96 plans — nearly one plan per creative, a template-testing day rather than a scale-up day.
| Date | New creatives | Plans | Read |
|---|---|---|---|
| 09-30 | 71 | 154 | Largest pulse, tied |
| 09-17 | 45 | 154 | Largest pulse, tied |
| 10-01 | 53 | 133 | Follow-through |
| 09-22 | 45 | 116 | Mid-window beat |
| 09-25 | 52 | 113 | Mid-window beat |
| 10-07 | 66 | 110 | Live at window close |
| 10-03 | 76 | 96 | Template-testing day |
Reinvestment stays disciplined: 322 of 2,930 plans (11.0%) point at creatives that ran in earlier windows. Most of the book is always fresh, but the AppLovin tier — the playables and the boss-fight video — is where the surviving assets live for months, not weeks.
What this data can and cannot tell you
Coverage shares are multi-attribution and overlap; exclusive is a targeting definition, not an outcome — a 752-plan Korean exclusive says where TREEPLLA points money, not what those plans earn. Estimated exposure is a global, lifetime figure per creative; it is neither a window increment nor a country-level number, and we use it as a scale signal only. Creative-to-country attribution beyond the exclusive definition is not observable at plan level. The two publisher listings are reported separately and never summed. All figures describe the 2026-09-09 to 2026-10-09 window as observed by CreativeScope on the retrieval date and will drift as records update.
Key takeaways
- The buy was still climbing at the end: 1,595 new creatives in the 61 days to October 9, the daily rate rising from 18.5 in August to 42.6 in early October, and the three heaviest days all falling in the final ten.
- Split the world into four buyers, not one: Korea (752 exclusive plans), the US (381) and Japan (380) each get their own book; Europe, Southeast Asia and Oceania share one flat 500–600-plan track with single-digit exclusives.
- Localization shows up in assets, not settings: the Korean lane runs its own subtitled video, four-plus Korean-copy images and a Korean-headline playable; Japan runs a separately titled video. The shared package stays English.
- Four lanes at once is a statement: Meta four-surface package (42.39% each) + Messenger, AdMob (21.81%), AppLovin (16.18%) and a Vungle/ironSource/Unity native tier, each lane carrying its own asset class — playables live only on AppLovin.
- Quality is concentrated where the exposure is: the window’s biggest single asset (1.48M est. exposure) is a 720p single-edit AppLovin video, and nine of the twelve hottest creatives are AppLovin assets.
Related reading
- Crownstone Survival Ads: 809 Plans That See Only Indonesia or the US — the isolation mirror image: two markets walled off, one Meta package carrying the buy.
- Tiles Survive! Ad Creatives: An ad themed around AI and strategic management — a 4X record that flattens 197 markets to one width and buys only two countries alone.
- Bulldozer Master Ads: Unlock Epic Bulldozers, Aimed at the US — the single-market isolation playbook at a twentieth of the scale, and the closest cousin to this game’s 752-plan Korean wall.
Want the raw structure behind this breakdown — every plan, creative and network for Idle Cat Gunner: Shooter RPG and 4,000+ other tracked games? Explore the data at creativescope.ai/mcp.
FAQ
What is Idle Cat Gunner: Shooter RPG?It is a side-view idle shooter for mobile by TREEPLLA, a Seoul-based studio operating under Neptune. Squads of cartoon cats auto-fire through biome maps while the player merges duplicate weapons into stronger guns; the store listing carries a 1M+ downloads label.
What does “exclusive plan” mean in this article?A plan whose target-country list contains exactly one market. Korea has 752 such plans, the US 381 and Japan 380. Exclusivity describes targeting decisions, not results — it says nothing directly about ROI or conversion in those markets.
Why do Audience Network, Facebook, Instagram and Threads all show 42.39%?They are one buy. Meta’s delivery system places the same package on all four surfaces, so each covers the same 1,242 plans. Messenger, at 32.46%, receives a subset. Coverage rates overlap and are not meant to sum to 100%.
How big is the playable investment?174 playable creative records carrying 233 plans in the window, effectively all delivered through AppLovin. The largest playable’s estimated exposure (1.42M) exceeds every Meta video in the sample, and identical-preview records have been live since mid-July.
Is this a new game?No. The store record carries no usable launch date, but the 1M+ downloads label, the 2,930-plan window and the standing playable register all point to a mature product in steady-state user acquisition, not a launch.
Why do some totals mention 1,399 creatives when the headline says 1,331?1,331 is the deduplicated count of distinct creatives. 1,399 counts the same asset once per network attribution; a video on three networks appears three times. Structural sections use the attribution count because that is how networks, formats and countries split.
Data notes
Source: CreativeScope ad-intelligence dataset (retrieved 2026-10-09), window 2026-09-09 to 2026-10-09, publisher listing TREEPLLA. “Coverage” is the share of the window’s 2,930 ad plans whose target-country or network list includes the item in question, computed with multi-attribution; overlapping shares therefore do not sum to 100%. “Exclusive” requires a plan’s target-country list to contain exactly one market. “Estimated exposure” is CreativeScope’s global lifetime estimate per creative, used here only as a relative scale signal. Creative, format and network counts in the structure section use network attribution (1,399 records / 2,615 attributed entries); the headline count of 1,331 is deduplicated. A second CreativeScope record for the same title under Neptune Company (1,038 creatives / 1,243 plans / 38.8M estimated exposure) is reported separately and excluded from every figure above.
The two-month trend chart plots the daily new-creative series for 2026-08-10 to 2026-10-09; a creative counts on the day it was first recorded, the orange line is a trailing seven-day average of that daily count, and those context totals are never mixed into the 31-day window figures. Creative descriptions in the spotlight section are based on frame-by-frame visual inspection of the delivered files. Live windows, family counts and exposure figures per spotlight come from the same dataset. Nothing in this article should be read as an endorsement of the targeting practices described; exclusive buying and market isolation are observed behaviors, not recommendations.
Get this live in your agent
These rankings come from a single MCP tool. Free key, ask in plain English.
Full genre × market grid?
Team plan unlocks every genre and market cut, plus AI teardowns.