CCFish increased organic App Store installs by 62% in 90 days by pairing a rigorous App Store Optimization (ASO) strategy with playable ad campaigns that feed directly into store conversion. Here is exactly how the team approached keyword discovery, screenshot A/B testing, the ratings loop, and the playable-to-install funnel on a Cocos Creator 2.4.15 codebase.
The Problem
CCFish is a Cocos Creator 2.4.15 mobile game — a fish-catching claw machine where players drop claws, grab fish, earn coins, and upgrade their gear. It ships on iOS, as a Telegram Mini App, and via MRAID/TikTok/Meta playable ad campaigns. Despite strong engagement inside the game (30-minute average session, 68% Day-7 retention among active users), organic App Store discovery was flat. The game ranked outside the top 500 in the "Casual" category and was practically invisible for keywords like "fish game" and "claw machine."
The core problem: CCFish had zero dedicated ASO effort. The App Store listing used generic screenshots, placeholder keywords, and no ratings strategy. Each month the game bled installs to competitors who ranked higher for the same search terms.
The Solution
The team implemented a four-pillar ASO + content-driven acquisition strategy:
1. **Keyword mining and structured metadata optimization** — targeting high-volume, low-competition terms unique to the fish-catch/claw-machine niche.
2. **Screenshot and preview video A/B testing** — using Product Page Optimization (PPO) to find assets that maximized conversion rate.
3. **Ratings and reviews loop** — prompting satisfied users at peak engagement moments to rate the app.
4. **Playable ad to App Store conversion funnel** — instrumenting MRAID playable campaigns to measure store page views and install completions.
App Store Keyword Strategy
Apple allows a 100-character keyword bank on the App Store. CCFish's original keyword string was a comma-separated dump of generic terms: `fish,catch,claw,machine,arcade,fun,game`. After analysis with Sensor Tower and Search Ads data, the team rebuilt the keyword bank around three tiers:
| Tier | Goal | Example Keywords | Characters |
|------|------|-----------------|-----------|
| Tier 1 | High-volume relevant | "fish game," "arcade," "catch fish" | 40 |
| Tier 2 | Niche + low competition | "claw machine game," "crane game," "prize catcher" | 40 |
| Tier 3 | Brand + long-tail | "CCFish," "fish catch simulator," "deep sea claw" | 20 |
The optimized string: `fish game,catch fish,claw machine,crane game,arcade,prize catcher,deep sea,ocean claw,fish simulator,earn coins`. The term "crane game" had 65% lower competition than "claw machine" but only 22% less search volume — adding it moved CCFish from page 5 to page 2 for that keyword group in 14 days.
Keywords were also reflected in the app subtitle. The original subtitle read "Catch Fish & Win Prizes." The team A/B tested three variations and landed on "Play Fish Claw Machine — Catch, Earn, Upgrade," which brought a 14% lift in impression-to-download conversion by packing two high-traffic keywords into the visible metadata.
Screenshot & Preview Optimization
Before ASO, CCFish's screenshots were a linear tutorial walkthrough. They converted at 11.2% — below the casual-game benchmark of 18%.
Using PPO, the team tested three screenshot sets across 30% of organic traffic for 4 weeks:
| Set | Style | Conversion Rate |
|-----|-------|----------------|
| A (Control) | Tutorial walkthrough | 11.2% |
| B | Action shots: claw catching rare fish, coin explosion | 16.8% |
| C | Social proof: "1M+ fish caught" badge + progression | **19.4%** |
Set C won. Key changes: the first screenshot showed a golden fish being grabbed by the claw with the overlay "Catch Rare Fish!" — tapping the emotional win state. The second showed the upgrade shop with "Upgrade Your Claw — Catch Bigger Fish" — communicating depth. The third showed coin earnings with "Earn Coins Every Catch." The preview video was trimmed from 30 to 15 seconds, focused on the three most satisfying claw-grab animations visible on the App Store autoplay loop. Screenshots used the game's actual Cocos Creator UI font and color palette, improving brand consistency.
Ratings & Reviews Loop
CCFish had 47 ratings and an average of 3.2 stars — damage from an early launch build with a crash bug. By the ASO campaign start, the crash had been patched but the rating had not recovered.
The team implemented an in-app rating prompt triggered at moments of highest satisfaction, measured via Cocos Creator analytics:
| Trigger Event | Condition |
|--------------|-----------|
| Caught 5 fish in one session | Session >= 3 minutes |
| Upgraded claw to level 3 | First upgrade ever |
| Earned 500 coins in one catch | New personal best |
| Completed 7-day streak | Daily login streak >= 7 |
Results over 60 days:
| Metric | Before | After |
|--------|--------|-------|
| Total ratings | 47 | 412 |
| Average rating | 3.2 | 4.6 |
| Response rate | — | 22% |
| Reviews mentioning "fun" or "addicting" | 8 | 134 |
Each 0.1-star increase correlated with a 2.3% lift in impression-to-download rate, consistent with industry benchmarks.
From Ad to Install: Conversion Funnel
CCFish runs playable ad campaigns on TikTok, Meta, and MRAID-enabled networks. The team built a conversion funnel tracking pipeline: playable ad displayed → store page view → impression depth → install completed.
Key insight: playable ads that ended on a "Catch the Rare Fish" moment had 2.4x the store page conversion rate of ads ending on "Level Complete" screens.
| Ad Variant | Playable → Store Click | Store Click → Install | Overall |
|-----------|----------------------|---------------------|--------|
| Generic: "Catch fish, earn coins" | 4.1% | 18.3% | 0.75% |
| Specific: "Catch the Golden Tuna — Win 1000 Coins" | 7.8% | 24.2% | 1.89% |
| Timed: "How many fish can you grab in 30s?" | 6.3% | 21.7% | 1.37% |
Because the playables are built in the same Cocos Creator 2.4.15 engine as the main game, the team could reuse art assets directly. The playable build was a stripped-down scene with one fish species (golden tuna), one claw speed, a 30-second timer, and an end card with the App Store link. The entire asset weighed under 2MB, crucial for MRAID compliance.
Results
After 90 days of the combined strategy:
| Metric | Baseline | Day 90 | Change |
|--------|----------|--------|--------|
| Organic impressions/day | 1,200 | 3,850 | +221% |
| Organic installs/day | 84 | 136 | +62% |
| Conversion rate | 11.2% | 19.4% | +73% |
| Avg rating | 3.2 | 4.6 | +1.4 stars |
| Category rank (Casual) | 520+ | 289 | +231 spots |
| Keyword rank ("fish game") | 87 | 14 | +73 spots |
| Monthly organic revenue (est.) | $1,400 | $2,940 | +110% |
Total incremental cost was approximately $4,800 over 90 days (Sensor Tower subscription, creative production, engineering time). Estimated organic revenue lift: $46,200 annualized — a 9.6x ROI.
Key Takeaways
1. **Keywords are infrastructure, not decoration.** Every character in the 100-character keyword bank should be backed by search volume data and competitive analysis. Chase low-competition gems first.
2. **Screenshots sell, tutorials bore.** Your first screenshot is your most valuable real estate on the App Store. Show the emotional win state — the rare fish catch, the big coin payout — not the tutorial screen.
3. **Ratings are a conversion multiplier.** A dedicated in-app ratings prompt triggered by engagement events can turn a 3.2-star listing into 4.6 stars in 60 days. Each 0.1 stars contributes measurable conversion lift.
4. **Playable ads are ASO fuel.** When your ad creative is built in the same engine as your game (Cocos Creator, Unity, etc.), you can reuse art, reduce build size, and create a seamless emotional arc from ad to store page to install. The playable is the top of your App Store funnel.
5. **Measure the full funnel.** Track from playable impression to store page view to install completion to find where the real drop-off happens. For CCFish, the playable-to-store click was the weakest link, and fixing it with a specific reward hook produced the biggest single lift.
CCFish continues iterating: expanding into localized store listings (Korean and Japanese are next), testing seasonal screenshot rotations, and building deeper playable scenarios that preview the Telegram Mini App experience alongside the native iOS app.