Allodial Games
Source reportReport

First mobile game

Build a 3D stuck-state hybridcasual puzzle.

Puzzle was the only big genre to grow in H1 2026, up ~20%, while total mobile game IAP fell 2%S1. At the sourced defaults, ads alone pay back a $0.50 install on day 75S4, S7.

See the ranking Run the payback
01

What should our first game be?

A one-mechanic 3D puzzle, monetized by ads plus light IAP, for three reasons.

A sort, screw or block-jam style game with premium 3D presentation, monetized by ads plus light IAP: boosters, no-ads, lives.

  1. It is where the growth is.

    Puzzle was the only big genre to grow in H1 2026, up ~20% to $8.07BS1. That is roughly $1.3B of new revenue (derived from those two figures), about 70% of all the revenue gained by genres that grewS1. The market still fell 2% overall because Strategy, RPG and Casino lost more than Puzzle gainedS1. Inside puzzle, the 2025 growth came from merge-2 (about 60% of the increase) and the new Sort, Screw and Block niches (about 40%)S4. Each of those niches doubled or betterS2, S4. Merge-2 is real growth, but it is IAP heavy and content heavy, so it is not a first title. The niches are.

  2. It is buildable by a small team.

    The core loop is one mechanic plus a level generator. No narrative, no multiplayer, no server-authoritative economy. 3D presentation is exactly where "good graphics" buys differentiation cheaply: one polished scene, good materials, satisfying physics and juiceJ.

  3. The unit economics can close on install ads.

    Sort puzzles buy installs at around $0.50 blendedS4. Ads start earning in the first session, and IAP layers on top. At the sourced defaults, ads alone pay back a $0.50 install on day 75S4, S7. That is the shortest payback path available to a first title.

What to avoid: classic match-3, merge-2 as a first title, 4X strategy, RPG, casino and shooters. The reasons are in the field table.

02

What does "easy to build" mean?

Easy means two to four people and a soft launch in two to four months.

Build tierTeamTime to soft launchWhat it means
Easy 2 to 4 2 to 4 months One core mechanic, procedural or hand-tuned levels, offline, no live content treadmill
Medium 5 to 10 6 to 12 months Meta layer (decorate, build, collect), meaningful content pipeline, LiveOps events
Hard 15+ 18+ months Online systems, deep economy, large art and content teams, tens of millions in UA to compete

Build tiers, team sizes and timings are our judgment.

03

How crowded is each category?

Big money is in the hard categories. The growth is in the easy ones.

Competition, Low / Med / High / Brutal, reflects revenue concentration and the success rate of new launches. Revenue is 2025 global IAP unless noted. Where two sources differ, the table gives the range.

Category Build 2025 revenue (IAP) Trend Competition Monetization First-title fit
Build first
Sort puzzleliquid, ball, goods sort Easy ~$280MS4 ($279MS2) +176% YoYS4 Med. Healthy middle class: 12+ games at $5M to $15M/yr. Leader (Magic Sort) ~18% shareS4 Hybrid, ads first Strong
Screw / pin puzzle Easy ~$177MS2 to $206MS4 ~2x YoYS2, S4 High. Leader (Screwdom) ~45% shareS4 Hybrid, friction-driven IAPS4 Strong
Block puzzleblock jam, color block Easy ~$183MS4 to $213MS2 10x+ YoYS2 High at the top. Color Block Jam ~75% of IAPS4; Block Blast earns an estimated $600K to $1M/day from ads aloneS6 Ads first, IAP layered Strong but crowded
Tile match / Match-3D Easy to Med not broken out Mature High Hybrid Good if differentiated visually
Idle / clicker / tycoon Easy to Med not broken out +40% YoY in H1 2025S3 Med Ads (rewarded) + IAP Good second option; strong 3D showcase
Not first
Solitaire / Mahjong Easy not broken out recently Stable High, older audience, ad-heavy Ads + IAP Moderate. Low creative headroom for graphics
Word games Easy not broken out DecliningS4 High Ads Weak. Legacy subgenre losing installsS4
Hypercasualrunner, one-tap Very easy ad-only Lost the most downloads in H1 2026S1 Brutal on volume Ads only Weak. Thin margins, short lifespans
Merge-2Gossip Harbor style Medium ~$1.4BS2 +80% YoYS2 High. Leader ~33% shareS2 IAP heavy Not first. Huge art and narrative content load
Farming sim Medium ~$955MS2 −2%S2 High. Township ~42%S2 IAP No
Time management Medium ~$135MS2 −12%S2 Med IAP No. Shrinking
Life sim Medium ~$135MS2 flat (+1.4%)S2 Med IAP No
Avoid
Match-3Royal Match style Hard in practice ~$4.8BS2 flatS2 Brutal. Dream Games holds ~30%; only 4 of 367 new launches passed $100K/monthS2 IAP Avoid
Slotscasino genre ~$6.9B, −7.6%S2 Medium build, hard business ~$2.7BS2 −14%S2 Brutal, D2C shiftS1, regulatory exposure IAP Avoid
Strategy (4X) Hard ~$9.3B (H1 2026)S1 −5%S1 Brutal IAP Avoid
RPG Hard ~$6.1B (H1 2026)S1 DecliningS1 Brutal IAP gacha Avoid
ShooterCoD Mobile tier Hard large n/a Brutal IAP Avoid

Build tier, competition and first-title fit are our judgment. Revenue figures are IAP only unless noted. Ad revenue (IAA) is not in most public subgenre totals, which understates the ad-driven genres this report favors. Superscripts link to the sources.

04

Where is the money, and where is the growth?

The big categories are flat or shrinking. The small ones are doubling.

Global mobile game IAP was ~$40B in H1 2026, down 2% YoY, with downloads down ~12%S1. Puzzle, at $8.07B (+20%), is closing on Strategy at $9.27B (down ~5%)S1. Casino as a genre was ~$6.9B in 2025 (−7.6%): slots ~$2.7B (−14%), coin looters ~$2.4B (−3%) and casino card games ~$784M (−3%)S2.

Subgenre 2025 IAP revenue Growth, year on year
Match-3 ~$4.8B flat
Slots ~$2.7B −14%
Merge-2 ~$1.4B +80%
Farming sim ~$955M −2%
Sort puzzle ~$280M +176%
Block puzzle ~$183M to $213M 10x+
Screw / pin puzzle ~$177M to $206M ~2x
Time management ~$135M −12%
Life sim ~$135M +1.4%
Nine subgenres with both a 2025 IAP revenue figure and a growth figure in the reportS2, S4; Strategy and RPG are left out because their revenue is for H1 2026. Slots is the largest casino subgenre. Revenue runs on one linear scale to $4.8B; where sources differ, the solid bar runs to the low figure and the lighter bar to the high one. Growth runs from −20% to +200%: "~2x" is drawn at +100%, "flat" as a square on zero, and Block's "10x+" is broken because it runs off the scale. Revenue is IAP only, so ad-driven puzzles are understated.
05

Can install ads pay for it?

CPIs are rising everywhere, which favors games that monetize from day one.

CPI benchmarks, 2026S5

GenreiOSAndroid
Hypercasual$1.50 to $2.50~$0.40
Casual / Puzzle~$3.00~$2.00
Simulation~$3.75~$2.50
Midcore~$4.50~$3.25
Strategy~$5.50~$4.00
Hardcore / RPG~$6.00~$4.50

Why $0.50 and $2 to $3 are not in conflict. Naavik's ~$0.50 for sort puzzles is a blended global figureS4, beside FoxData's all-genre global blend of $0.56S5. FoxData does not state the country mix behind its per-platform genre figures; read against that blend, they behave like Tier 1 weighted figuresJ. Treat $0.50 as what a globally bought sort install costs and $2 to $3 as the Tier 1 reference. Never pair a global CPI with Tier 1 revenue, or the reverse.

CPI by region, all genresS5

~$1.68

North America, blended.

All genres · 2026

~$0.53

Europe, blended.

All genres · 2026

~$0.14

LATAM, blended.

All genres · 2026

Global blended CPI is ~$0.56, up 30% YoYS5.

Retention benchmarks

BenchmarkD1D7D30Source
All mobile games, median~23%~4%under 1%S8
All mobile games, top 25%~30%6 to 7%1.6 to 1.8%S8
All mobile games, top 10%~40%11 to 12%n/aS8
Global average, games~27%n/an/aS5
Competitive target30%+15 to 18%n/aS5
Hybridcasual, typicaln/a~20%~10%Gamigion hybridcasual overview, Q1 2025HC
Sort puzzlesn/an/a~8%S4
Screw (Screwdom)n/an/a~17%S4
Block (outside leaders)n/an/a~4%S4
Block Blast, the ceiling51%32%21%S6

Why our soft-launch gate is D1 35%, above the 30% target

  1. 30% is the market line.

    It is the "competitive positioning" targetS5, and it equals the top quartile of all gamesS8.

  2. 35% is our own gate.

    It sits between the top quartile (~30%) and the top decile (~40%)S8, because we only gate before spending real money on paid installsJ.

  3. D7 15% matches the low end of the competitive target.

    That is the low end of 15% to 18%S5, and below typical hybridcasual at ~20%HC.

Read the retention table honestly. The niche figures are for games that made it. Across all launches, the median D30 is under 1%S8. The payback defaults assume we hit sort-niche retention. The soft-launch gate is what tests whether we do.

Why the hybridcasual puzzle wins the ROI testJ

  1. Low CPI pool.

    Casual puzzle traffic is cheap in Android and Tier 2/3 geos, so you can buy enough installs to learn without large spend.

  2. Day-zero revenue.

    Interstitials and rewarded video start paying in the first session. IAP-only genres need weeks to recover CPI.

  3. Testable before building.

    Run CPI tests on ad creatives (video of the mechanic) before writing the full game. If the creative cannot get a cheap install, the game will not either.

  4. Small surface to tune.

    One mechanic means difficulty curve, ad placement and booster pricing are the main levers. Fast iteration loops.

06

How long until an install pays for itself?

Scale only when the curve projects payback inside ~90 days.

Set what an install costs, what a daily player earns from ads and IAP, and how many players come back. The result shows the day cumulative revenue per install first covers its cost. Every default is sourced, derived from a source, or our own soft-launch gate.

CPI preset

Sort puzzles run around $0.50 blended globalS4; casual and puzzle benchmarks are ~$2.00 on Android and ~$3.00 on iOS, Tier 1 weightedS5.

IAP preset

A ceiling reference, not a target: AppsFlyer casual D90 IAP of $1.34 per installS7 over 8.49 player-days. The casual average includes IAP-heavy match-3 and merge.
Cost of one install.
Ad revenue per daily active user. Default derived from AppsFlyerS7.
IAP revenue per daily active user. Default is ads only.
Retention
Our soft-launch gate.
Our soft-launch gate. Shown against the fitted curve; not used in the fit.
Sort puzzle D30S4. Held at or below D1.

Days to payback

Day 75

A $0.50 install earns $0.55 by day 90, 110% of its cost.

Model: retention is a power curve r(t) = a · t−b fitted through D1 and D30, with day 0 at 100%. Lifetime value LTV(t) = ARPDAU × the sum of r(d) for d = 0 to t, where ARPDAU is ad plus IAP revenue per daily active user. The ad default is derived: AppsFlyer's global D90 ad revenue for casual games, $0.55 per installS7, divided by the 8.49 player-days the default curve yields through day 90.

  • LTV at D30$0.32
  • LTV at D90$0.55
  • LTV at D365$1.17
  • D7, entered versus fitted15.0% vs 15.0%
  • Curve fita = 0.350, b = 0.434

What the defaults say

  1. At the sort-puzzle CPI, ads alone pay back inside 90 days.

    A $0.50 install pays back on day 75 on ads alone, and on day 6 with casual-average IAPS4, S7.

  2. At Tier 1 casual CPIs, the game only pays back with real IAP.

    Ads alone do not pay back a $2.00 or $3.00 install inside 365 days. With casual-average IAP they pay back on day 100 and day 214S5, S7.

  3. The other lever is retention.

    At Block Blast's retention (D1 51%, D7 32%, D30 21%) and its own implied ad revenue per playerS6, ads alone pay a $0.50 install back in 22 to 47 days. That ad figure is $600K to $1M a day over ~15M DAU, so $0.04 to $0.067 per DAU (derived).

07

Which version do we build?

Screw first, sort second, block third.

3D screw or pin physics puzzle with a fresh theme

Why
Strongest D30 retention of the new niches (~17% for the leader)S4, so LTV holds up. Physics plus 3D camera rotation is a natural place for premium visuals and tactile satisfactionJ. Market roughly doubled in 2025S2, S4.
Risk
Leader holds ~45%S4. Differentiate on presentation and theme (materials, destruction, a setting with personality), not mechanic.
Build
3 to 4 months to soft launch, 3 people (gameplay engineer, 3D artist, designer/UA)J.

Sort puzzle with a premium 3D presentation

Why
Least concentrated of the three niches. A dozen titles earn $5M to $15M a yearS4, which proves a mid-tier hit is reachable without dethroning the leader. CPI around $0.50S4.
Risk
Lowest-effort genre means the most clones. Visual quality and level design are the moatJ.
Build
2 to 3 months, 2 to 3 peopleJ.

Block jam / color block puzzle

Why
Biggest proven ceiling: Block Blast at an estimated $600K to $1M/day in ads, ~15M DAUS6. Fastest growth curveS2.
Risk
Most crowded at the top, and Color Block Jam owns most of the IAPS4. Pursue only with a distinct twist.
Build
2 to 3 months, 2 to 3 peopleJ.

Honorable mention3D idle tycoon. A strong visual showcase, and growing (+40% YoY in H1 2025S3), with rewarded ads as the core monetization. It has a longer tuning tail on the economy than a puzzle, which is why it ranks fourthJ.

08

What do we do first?

Test the ad creative before writing the game.J

  1. Creative test first (2 to 3 weeks).

    Make 3 to 5 short ad videos per candidate concept (screw, sort, block) and run small CPI tests on Meta and AppLovin in a low-cost geo. Pick the concept with the best IPM and CPI.

  2. Prototype the winner (6 to 8 weeks).

    50 levels, ads integrated, analytics (D1/D7, ad impressions per DAU, level funnel).

  3. Soft launch in a low-CPI geo.

    Gate on D1 above 35% and D7 above 15% before spending in the US.

  4. Scale only when the ROAS curve projects payback inside ~90 days.

    Put soft-launch retention and revenue per DAU into the payback calculator in place of the defaults.

Sources

Where do these figures come from?

Every figure on this page comes from one report and its sources.

The report: First Mobile Game: Genre Shortlist and Recommendation, revision 3, 4 October 2026. Market figures cover 2025 to H1 2026.