Castle Busters Ads: Siege Footage Built for the Feed
Castle sieges cut to fit the feed: how a Voodoo voxel-action title ran ~1,100 ad plans a day through October, let Turkey lead, and peaked at 1,719 in Halloween week.
Emma Lee
Editorial Research Contributor
Published
Updated
Castle Busters Ads: Siege Footage Built for the Feed
Castle sieges cut to fit the feed: Castle Busters kept roughly 1,100 ad plans a day in delivery through October, fed fresh creative into Halloween week until the count peaked at 1,719 on October 30, and let one evergreen concept run for three months.
What this breakdown covers
- A month of steady output, then one loud week. 1,774 creatives and 6,324 ad plans went out across October, with 353 of the creatives carrying a reinvestment flag. Daily in-delivery plans hovered near 1,100, dipped to 896 on October 15, then climbed through Halloween week to a peak of 1,719 on October 30.
- Türkiye led the book. 1,927 plans — 30.5% coverage — targeted Türkiye, including 146 plans that target it alone. No other market came close on exclusives: France had 17, Spain 15.
- Europe was the belt, America was the sideline. Twenty European markets sat between 21% and 30% coverage; Mexico 22.4%, Great Britain and Germany 21%. The United States ran at 14.1% (891 plans), roughly thirty-third among 209 observed markets — for a Voodoo title, an unusually quiet home stretch for the Anglosphere.
- Meta four carried the volume, TikTok took a double-digit lane. Facebook and Instagram each covered 61.9% of plans, Audience Network and Threads 53.2%. TikTok sat at 18.5% — a share casual merge books rarely touch, and one that fits voxel-destruction footage cut for fifteen-second feeds.
- Google's share was small; its units were not. AdMob and YouTube together covered under 5% of plans, yet the two heaviest creatives in the set — 67.6 million and 40.7 million in lifecycle exposure — both ran on the Google line, each with dozens of sibling variants.
- WhatsApp: zero. The same Meta stack that carried the volume left WhatsApp entirely untouched in October — the opposite of the preheat books that use it as a cheap social lane.
The shape of the month: a production line with one loud week
Castle Busters is an Aarrow-era Voodoo action title — voxel units, castle walls, base-rush battles — that has been live for over a year, and its October book reads like a factory schedule rather than a campaign. Every day of the month had between roughly 900 and 1,450 plans in delivery, and the daily count barely moved for three weeks: 1,125 on October 2, 1,207 on October 9, 1,151 on October 20, all within the same band. The low point, 896 plans on October 15, is the only dip worth naming.
Then the last week of October broke the band. From October 26 the count climbed — 1,351, 1,427, 1,351, 1,449 — and on October 30 it hit 1,719, the highest day of the month, with 1,651 still in delivery on the 31st. New creatives arrived with the same rhythm: the freshest families landed October 26 through 30, which means the surge was not old inventory being pushed harder but new supply being fed in. Whatever the internal reason — a seasonal beat, a store feature, a quarterly target — the delivery data shows a book that was deliberately wound up in the final week rather than one that simply drifted upward.
Türkiye leads, Europe is the belt, America is thirty-third
The market structure is the most distinctive thing in this book. Türkiye tops it at 30.5% coverage with 1,927 plans — and, more tellingly, 146 plans that target Türkiye alone. Market-exclusive plans are a deliberate shape: they carry creative or bidding logic that only makes sense in that market, and no other country in this book comes close. France holds 17 exclusives, Spain 15, Switzerland 9; everything else is single digits.
Behind Türkiye stretches a European belt: France, Italy, Switzerland, the Netherlands and Spain between 27% and 28%, Scandinavia and the eastern flank — Sweden, Norway, Romania, Hungary, Greece, Poland, Austria, Belgium, Ireland, Finland, Portugal, Denmark, Czechia — from 23% to 26%. Mexico at 22.4%, Great Britain and Germany at 21%. That is twenty markets covered in one band, which is what a belt looks like: no single European market is special, all of them are served.
The United States is the exception that explains the strategy. At 14.1% coverage with 891 plans — roughly thirty-third of the 209 markets observed — America took a smaller share of this book than Slovenia or Latvia. For a French publisher whose home continent prices inventory more cheaply than the US and consumes mid-core action creatives readily, October was a month of buying Europe deeply and letting America wait. Whether that is a deliberate rotation or a persistent home-market habit, the shape is unambiguous: this was not a US-first month.
Meta four for volume, TikTok for a lane, Google for weight
The channel stack splits cleanly into three jobs. Facebook and Instagram each covered 61.9% of October plans, Audience Network and Threads 53.2% — the Meta four moving together as one fanned-out plan, exactly the pattern a high-volume action book needs for reach. TikTok is the second story: 18.5% of plans, 1,171 of them, a share that merge and puzzle books almost never take and that this category — voxel destruction, castle sieges, unit swarms — wears naturally in a fifteen-second feed. AppLovin at 6.2% and Unity at 5.4% filled the network floor; Messenger sat at 4.3%.
The third job is Google's, and it is a weight-bearing one. AdMob covered 4.1% of plans and YouTube 0.4%, but the two heaviest creatives in the entire October set both ran on that line: a video family in delivery from September 15 to 24 carrying an estimated 67.6 million in lifecycle exposure across 46 sibling variants, and a second family from September 2 into October with 40.7 million across 39 variants. A third Google unit, first delivered October 12 under Voodoo's own naming convention — CCA_FC53_1080x1920_WW — added 12.0 million and was still in delivery on November 13. Share of plans measures how widely a network is used; exposure measures how much a network's units weigh. In October those two measures pointed at different networks.
The Meta book worked differently: smaller units, huge families. Its heaviest October creatives carried 1.4 million, 1.0 million and 0.8 million in lifecycle exposure — two orders of magnitude below the Google flagship — but each sat inside a family of 40 to 80 near-identical siblings, almost all flagged as reinvestment. That is a different production model: Meta buys frequency through replication, Google buys weight through concentration, and Castle Busters ran both at once.
What to copy, and what to doubt
Three things here are portable. One: the belt structure — twenty markets bought within a seven-point band, with a single market carrying the exclusives — is a cheaper way to hold Europe than bidding the US first, and the 146 Türkiye-only plans show how a leader market gets its own bidding logic without dragging the belt. Two: the dual production model, where Meta replicates one concept into dozens of siblings while Google concentrates weight into few units, is a division of labour a UA lead can copy explicitly rather than by accident. Three: winding the whole book up for one week, with fresh creatives timed to the same days as the plan peak, is a repeatable cadence for any seasonal or commercial beat.
What to doubt is the quiet part. This teardown sees delivery, not outcomes: no install-by-market, retention or revenue series is cited, and the 3.38 million Google Play net increment covers Android only. Estimated exposure is a lifecycle weight per creative, useful for comparing creative mass and nothing else. And the Halloween-week climb is an observation about timing, not a claim about cause — the data cannot say whether the October 30 peak answered a seasonal audience, a store feature, or an internal target. It can only say the machines were turned up, on purpose, with new supply.
Want to see which markets get their own exclusive plans — and which networks carry the weight vs the volume?
Check CreativeScope plansQuestions we get about this one
- What is Castle Busters?
- A voxel-art action title from Voodoo — castle sieges, unit swarms, base-rush battles — live since July 2025 with a public store download label above 10 million. This breakdown covers its October 2026 ad delivery only.
- How big was the October book?
- 1,774 creatives and 6,324 ad plans were in delivery across October 1 – 31, 2026, with 353 creatives carrying a reinvestment flag. Daily in-delivery plans ran near 1,100 for most of the month, dipped to 896 on October 15, and peaked at 1,719 on October 30.
- Which markets carried the buy?
- Türkiye led at 30.5% coverage with 1,927 plans, including 146 that target it alone — the largest exclusive block by far. Twenty European markets followed in a 21–28% band; Mexico ran at 22.4% and Great Britain and Germany at 21%. The United States took 14.1% (891 plans), roughly thirty-third of 209 observed markets.
- Which channels did it use?
- Facebook and Instagram each covered 61.9% of plans, Audience Network and Threads 53.2%, TikTok 18.5%, AppLovin 6.2% and Unity 5.4%. AdMob (4.1%) and YouTube (0.4%) covered few plans but carried the two heaviest creatives — 67.6 million and 40.7 million in lifecycle exposure. WhatsApp was untouched in October.
- Why does TikTok matter here?
- Because 18.5% of plans is a double-digit share casual and puzzle books almost never take. Voxel-destruction gameplay footage cuts naturally into short vertical feeds, and the share suggests Voodoo treats TikTok as a standing lane for this title, not an experiment.
- Did Halloween drive the late-month surge?
- The timing matches: the plan count climbed through the last week of October to its 1,719 peak on October 30, and the largest new-material days — 85 on October 29, 71 on October 26 — sat inside the same stretch. The data shows the wind-up and its timing; it cannot show the motive, so we describe the surge as deliberate, not seasonal by proof.
Method. Figures come from CreativeScope's delivery data for Castle Busters (VOODOO) over October 1 – 31, 2026, with latest material snapshots dated November 13, 2026 and latest ad snapshot month October 2026. Creative counts are de-duplicated per game and creative; plan counts per game and plan. Daily in-delivery counts are per Asia/Shanghai natural day. Google Play download net increment (3,377,492) covers Android only, October 1 – 31. Public store download label (10 million+) and release timing come from the store listing.
Scope limits. Channel and market shares are computed on plan coverage with multi-attribution, so they overlap and do not sum to 100%. Estimated exposure is a global lifecycle estimate per creative, used only to compare creative weight — it is not a period increment or a view count. Market-exclusive counts refer to plans whose targeting includes that market alone. No install, retention or revenue figures are cited, and no causal claim is made between delivery timing and install timing.
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.
