Ad Creative Volume vs Reuse: What Six Titles Actually Did
Six top mobile titles shipped 628 to 9,360 ad creatives in one 30-day window. The reuse column splits them wider: 2.2% to 41.0%. Two machines, one market.
Daniel Brooks
Creative Strategy Analyst
Published
Updated
AI has made ad creative cheap to produce, and every UA feed now tells you to test more. The tracked record suggests the real dividing line runs somewhere else: how much of your creative library you keep shipping after the test. One 30-day window, six top titles, two clearly different operating systems.
Key takeaways
- Creative volume and creative reuse behave like two different operating systems. The highest-volume title in the window reinvested only 2.2% of its library; the lowest-volume 4X entrant reinvested 41.0%.
- Plans per creative separates them further: depth-led titles ran 3.0 to 4.6 ad plans behind each creative, while volume-led titles ran roughly 1.2.
- Inside one genre - idle mining - two titles with almost identical creative counts ran completely different structures. A follower does not have to copy the leader's machine.
- Similar-creative clusters on head creatives give a rough read on how heavily a motif is being copied across the market.
- AI lowers the cost of volume. It does not automatically build the reuse engine, because reuse is an asset-compounding decision, not a production decision.
Data scope. CreativeScope tracked ad records, August 29 to September 28, 2026, Asia/Shanghai. One creative = one game-material record counted once; one ad plan = one game-plan record counted once. Country and platform shares reflect overlapping plan-target coverage and do not sum to 100%. Exposure figures are lifetime estimates for creatives active in the window and serve as a delivery-heat reference only. All figures in this article come from the same single query pass against the same six game records.
Not covered. Spend, CPI, ROAS, conversion and causality are outside what tracked records can support. Plan counts are not budget. Where a download count appears, it is a store-download snapshot, not an outcome claim.
What's inside
The question the feed keeps asking
Open any UA newsletter this quarter and you will find the same advice: AI now generates creative at near-zero marginal cost, so the winning move is to test more concepts, faster. That advice is not wrong. It is just incomplete, and the incomplete part is expensive.
What the advice rarely tells you is what "more" means relative to anyone else. Volume advice without a reference point is a treadmill. The tracked record offers one: within a single 30-day window, six prominent titles put between 628 and 9,360 distinct creatives into delivery. That is a 15-times spread at the top of the market, before you even reach the long tail.
The more useful question is not "how many concepts should we test" but "what does our tested library do afterward". That is where the record splits.
Two operating systems, one market
Ranked by tracked creatives in the window, the six titles form two clusters. The first cluster ships thousands of creatives and reinvests in almost none of them. The second ships hundreds and reinvests in a third or more.
| Title | Genre | Creatives | Ad plans | Plans / creative | Reuse share |
|---|---|---|---|---|---|
 Royal Match volume-led | Match-3 | 9,360 | 12,340 | 1.32 | 2.2% |
 Gossip Harbor volume-led | Merge / story | 5,885 | 7,029 | 1.19 | 6.8% |
 MONOPOLY GO! hybrid | Board | 5,218 | 18,443 | 3.53 | 26.0% |
 Top Force depth-led | 4X | 946 | 4,305 | 4.55 | 41.0% |
 Gold and Goblins volume-led | Idle mining | 675 | 932 | 1.38 | 11.4% |
 Idle Zombie Miner depth-led | Idle mining | 628 | 1,893 | 3.01 | 38.4% |
Reuse share = reinvested creatives / total tracked creatives in the window. Plans per creative = ad plans / creatives. Both derive from the same single-window query; overlapping plan-target coverage means shares do not sum to 100%.
The two ratios tell one story from two ends. The volume league - Royal Match, Gossip Harbor and Gold and Goblins - runs roughly 1.2 to 1.4 plans behind each creative and keeps almost none of the library under reinvestment. The depth league - Top Force and Idle Zombie Miner - runs 3.0 to 4.6 plans behind each creative and keeps 38% to 41% of a much smaller library in active rotation.
Bars scaled within each chart; the two charts are not on a shared scale. Values from the same query pass as the table above.
MONOPOLY GO! sits deliberately between the two systems. It ships volume-league creative counts while running depth-league plans-per-creative, and it reinvests in about a quarter of its library. Licensed-IP titles can do this because a single licensed asset carries recognition across many plan variations at once; the record shows its United States plans alone included more than 2,600 country-exclusive records in the window.
What the record shows inside the creatives
Aggregate ratios become concrete when you look at individual head creatives and their similarity clusters. CreativeScope tracks, for each creative, how many similar creatives exist across the market - a rough proxy for how heavily a motif is being copied.
- Royal Match's highest-exposure AdMob creative in the window carried 101 similar creatives and a delivery window of days. In a library of 9,360 that individual creative is disposable; the system is built so it can be.
- Gold and Goblins' top window creatives carried similarity clusters of 170 to 192. The mining-coins motif has been copied across the market so widely that any single instance is one drop in a large pool.
- Idle Zombie Miner's head creatives carried clusters of 25 to 53 - roughly a quarter of the motif leader's cluster sizes. Its weight-counting haul hook is recognizably related, but not a direct clone.
- Top Force's highest-heat creative in the window sat in a cluster of one, and ran for roughly one hundred days. The title's depth structure depends on a small number of long-lived, hard-to-copy executions rather than a flood of near-identical variants.
Read together, the cluster sizes sketch the copying cost of each system. Volume-led libraries tolerate high similarity because they replace creatives continuously. Depth-led libraries need executions that competitors cannot cheaply imitate, because each one carries months of delivery.
The follower test: one genre, two machines
Genre comparisons sharpen the point. Gold and Goblins (launched September 2020) and Idle Zombie Miner (launched September 2023) both sit in idle mining, and their window creative counts are nearly identical - 675 versus 628. If creative count were the story, these two would be twins. The record says otherwise.
- Gold and Goblins ran 932 ad plans across a wide network spread - its largest single-platform share was AdMob at 51.4% of plan coverage, with meaningful volume on Facebook, AppLovin and Unity besides.
- Idle Zombie Miner ran 1,893 plans - roughly double - concentrated on Meta placements at 81.4% plan coverage plus AdMob at 14.2%.
- Their reuse structures differ just as sharply: 11.4% versus 38.4%.
Two products in the same genre, with the same size creative library, running twice-different plan structures and three-and-a-half-times-different reuse rates. The genre did not decide the machine. The operator did.
This matters for anyone benchmarking against a category leader. The record does not support the simple reading that followers decline in step with the leader they copy. What it does support is narrower and more actionable: within a genre, the creative-system choice - volume or reuse - is a real strategic variable, visible in the numbers, and not dictated by the genre itself.
Where AI actually fits
AI changes the cost side of the volume equation, and the record is consistent with that: the volume-league titles sustain their libraries only through continuous, high-frequency replacement - Royal Match recorded single days in the window with more than 600 new creatives. Producing at that pace without AI assistance is no longer realistic for most teams.
But the depth league shows what AI does not do by itself. Top Force and Idle Zombie Miner keep 38% to 41% of much smaller libraries under reinvestment. Reuse is a portfolio decision: which executions earn a second month of delivery, which hooks get rebuilt rather than retired, which motifs stay in rotation after the test read is in. That decision does not get easier when generation gets cheaper; if anything, cheaper generation makes the selection discipline more valuable, because the discard pile grows faster than the keep pile.
A practical read for your own pipeline: pull your last 30 days of tracked creatives and compute the same two numbers - volume and reuse share. Then place yourself honestly. A high-volume, low-reuse profile is a legitimate system when your category rewards freshness. A low-volume, high-reuse profile is legitimate when your executions are hard to copy. The profile that is hard to defend is the accidental one: high volume, no reuse discipline, and no memory of why any test won.
Creative spotlights
Four head creatives from the window, one per system, with the frame-level detail behind the observations above.
Spotlight 1 - Royal Match: the disposable head creative
Hook
At 0:00 a crowned king in purple robes stands on castle-brick steps beside a spike trap, and a full match-3 board fills the lower two-thirds of the frame. By 0:02 he lifts a translucent teal shield as a pebble flow begins pouring from the shaft above. The promise is instant: hold the line, or the board below gets crushed.
Body
From 0:05 to 0:40 the flow thickens and the board pays the price. The king crouches lower under the shield, a white cartoon hand taps tiles near the breach, and the flow carves a jagged notch into the right side of the board. A stamina button shifts from green to orange as the hold drags on.
End sequence
At 0:55 the ad closes on the game logo, six thumbnail panels of board scenes - snake, waterfall, green ooze, lava - and a green play-now button.
Takeaway
A hundred and one similar creatives tracked market-wide, a delivery window of days, and a library of 9,360 behind it. The head creative is designed to be replaceable - that is the system working, not failing.
Spotlight 2 - Top Force: the long-lived depth execution
Hook
At 0:00 a top-down snowy battlefield spreads under a red supply road: red camps numbered 2,000 face green camps numbered 1,040 to 1,536, and a walled ice base with a helicopter pad sits at the bottom of the frame. Every unit count is on screen from the first second.
Body
Between 0:05 and 0:18 the camera tracks the road as camps merge and swell - a green camp climbs to 4,000 with a flash, the base glows teal, and red 4,000 camps answer from the ridge. The counting is the content: the ad is legible without a single line of dialogue.
End sequence
By 0:26 the field holds three 4,000-class camps - red, yellow and green - and the ad closes on the standoff rather than a store card, with the copy family ("build a military base on the front lines, arm your soldiers") doing the conversion work.
Takeaway
A similarity cluster of one means competitors have not cheaply copied the execution. Combine that with 4.55 plans per creative and a 41% reuse share, and the title's structure is legible: few concepts, long lives, heavy plan coverage behind each one.
Spotlight 3 - Gold and Goblins: the widely copied motif
Hook
At 0:00 the level opens on a dirt mine floor carpeted with gold coins and gem shards, an orange counter reading 100 in the corner, and treasure chests half-buried in the debris. The scene is pure extraction from the first frame.
Body
At 0:14 two goblin miners dig at a dirt notch as the counter climbs to 288. By 0:25 a white glove points at a fresh cut and coins and magenta crystals pour out while the counter reads 723 - the digging is the dopamine, and the counter keeps the score.
End sequence
At 0:34 a waterfall of coins pours into a wooden barrel, and the ad closes on a download prompt with the event copy: "Special Idle Events to Earn Valuable Rewards and Grow Your Mining Operation Even Faster".
Takeaway
This execution ran for more than eight months with a cluster of one. Its sibling head creatives, carrying the same motif, sit in clusters of 170 to 192. The motif is market property now; the long-running execution is the defensible part.
Spotlight 4 - Idle Zombie Miner: the follower with its own machine
Hook
At 0:00 a mine cart waits beside a glowing green ore field, a black tag reading 0 lb, a wooden bridge crossing a blue gem wall. A purple 900 sits in the corner - the target is already on screen.
Body
At 0:15 the cart reads 10 lb and a ladder of tags - 10, 20, 100, 200, 500 lb - lays out the hauls to come. By 0:30 gold bricks spill from the ore field at 11 lb, and by the final seconds the counter reads 29 lb on the same route.
End sequence
The ad never leaves the bridge. The loop - load, haul, weigh, return - carries the full runtime.
Takeaway
A related-but-not-cloned hook (clusters of 25 to 53 versus the motif leader's 170 to 192), a Meta-concentrated plan structure, double the plan count of its genre peer, and a reuse share near the depth league. The record's clearest argument that a follower can run a different machine.
CreativeScope take
Volume and reuse are not rungs on a maturity ladder. They are two machines, and the tracked record lets you tell them apart with two numbers. The six titles here shipped between 628 and 9,360 creatives in one window, and the spread tells you less than the reuse column does: 2.2% to 41.0%. When AI makes volume cheap, the durable edge migrates to the decisions AI does not make - which executions earn a second month, which hooks get rebuilt, and how much plan weight sits behind each survivor.
Want to run the same two numbers on your category? CreativeScope tracks creatives, ad plans and reuse behavior across 15 networks and 53 markets.
FAQ
How many ad creatives should a mobile game test per month?
There is no single right number. In one 30-day window, six prominent titles shipped between 628 and 9,360 tracked creatives. What the record shows is that volume only means something next to its pair: how much of the tested library stays in delivery afterward. High volume with near-zero reuse and high volume with deep reuse are both real systems - but they are different businesses.
What is a creative reuse rate and how is it calculated?
In this article it is the share of a title's tracked creatives in the window that are under reinvestment - still backed by new ad plans rather than retired after the test. Divide reinvested creatives by total tracked creatives in the same window. The six titles here ranged from 2.2% to 41.0%.
Does higher creative volume guarantee better user acquisition performance?
No, and tracked ad records cannot answer performance questions - they do not carry spend, conversion or retention data. What the record does show is structure: which titles run volume-led versus depth-led systems, and how differently two same-genre products can operate at identical creative counts.
Why did the two idle mining games run such different structures?
Gold and Goblins and Idle Zombie Miner shipped nearly the same number of creatives (675 vs 628) in the window, but one ran twice the ad plans with a 38.4% reuse share and a Meta-concentrated plan structure, while the other spread plans across more networks with an 11.4% reuse share. The genre did not dictate the machine; the operator's choices did.
Where does this data come from and what are its limits?
From CreativeScope's tracked ad records for August 29 to September 28, 2026, queried in a single pass. Country and platform coverage overlap and do not sum to 100%; exposure is a lifetime estimate used as a delivery-heat reference; plan counts are not spend. See the data notes below for the full scope.
Data notes. All figures come from CreativeScope tracked ad records for Aug 29 to Sep 28, 2026: one creative = one game-material record, one ad plan = one game-plan record; country and platform shares reflect overlapping coverage and do not sum to 100%; exposure and similar-creative counts are estimates, not performance results. The cover is an AI-generated original with no text; videos, posters and table icons come from platform stores and belong to their respective advertisers and developers.
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.