10 Strategies to Lower Your CPI While Maintaining Quality
14 min read • Updated Jan 2024
→
10 Strategies to Lower Your CPI While Maintaining Quality
By Cubemobi Optimization Team • 14 min read • Updated January 2024
Reducing your Cost Per Install (CPI) without sacrificing user quality is the holy grail of
mobile user acquisition. Here are proven strategies that our top advertisers use to achieve
better efficiency.
1. Expand Your Creative Library
Creative fatigue is one of the biggest drivers of increasing CPIs. Users who have seen your
ad multiple times become blind to it, reducing CTR and conversion rates.
- Maintain at least 5-10 active creatives per campaign
- Refresh creatives every 2-4 weeks
- Test different concepts: gameplay, story, humor, ASMR
- A/B test thumbnails for video ads
2. Optimize for Post-Install Events
Instead of optimizing purely for installs, set up campaigns to optimize for in-app events
like registration or first purchase. This often results in:
- Higher quality users with better LTV
- Lower overall acquisition costs when measured against revenue
- Better retention rates
3. Leverage Dayparting
Not all hours deliver equal value. Analyze your conversion data by hour of day:
- Identify peak performance hours (often evenings and weekends)
- Pause or reduce bids during low-converting hours
- Consider timezone differences for each target country
4. Test New GEOs
While Tier 1 markets offer scale, Tier 2 and Tier 3 markets can offer excellent value:
- CPIs in Southeast Asia can be 50-70% lower than the US
- Some emerging markets have strong payer populations
- Less competition means better inventory access
5. Implement Blacklists and Whitelists
Use performance data to curate your traffic sources:
- Blacklist sources with consistently poor retention or fraud signals
- Whitelist top-performing placements for dedicated campaigns
- Review and update lists weekly
6. Optimize App Store Presence
Your app store page is the final step in the conversion funnel:
- A/B test app icons and screenshots
- Optimize descriptions with relevant keywords
- Respond to reviews to improve ratings
- Ensure app size isn't a barrier (especially for cellular downloads)
7. Use Rewarded Video Strategically
Rewarded video often delivers higher-quality users at competitive CPIs:
- Users actively choose to engage with the ad
- Higher completion rates lead to better brand impression
- Consider the reward type in apps where you advertise
8. Implement Smart Bidding
Use our automated bidding features:
- Target CPI bidding adjusts bids automatically to hit your goals
- ML algorithms identify optimal bid levels by segment
- Reduces manual work while improving efficiency
9. Segment and Personalize
Create separate campaigns for different user segments:
- Segment by device type (phone vs. tablet)
- Segment by OS version (newer devices often have higher LTV)
- Use different messaging for different demographics
10. Monitor and Act on Fraud
Fraud inflates CPIs by wasting budget on fake installs:
- Review MMP fraud reports regularly
- Set strict time-to-install thresholds
- Block sources with suspicious patterns
- Claim refunds for confirmed fraudulent traffic
Remember: CPI reduction should never come at the expense of user quality.
Always monitor retention, engagement, and revenue metrics alongside CPI. A $1 install that
never opens the app is worse than a $3 install that becomes a paying customer.
← Back to article list
Understanding and Improving ROAS
11 min read • Updated Jan 2024
→
Understanding and Improving ROAS
By Cubemobi Analytics Team • 11 min read • Updated January 2024
Return on Ad Spend (ROAS) is the ultimate measure of your user acquisition success. It tells
you how much revenue you generate for every dollar spent on advertising.
Calculating ROAS
The basic formula is simple:
ROAS = Revenue from Acquired Users / Ad Spend × 100%
For example, if you spend $10,000 on ads and acquired users generate $15,000 in revenue, your
ROAS is 150%.
ROAS Timeframes
ROAS is typically measured at specific points in time:
- D0 ROAS: Revenue on install day (useful for apps with immediate
monetization)
- D7 ROAS: Revenue after one week (common benchmark for games)
- D30 ROAS: Revenue after one month (better LTV indicator)
- D90/D180 ROAS: Long-term value measurement
What's a Good ROAS?
Target ROAS varies by business model and payback period tolerance:
- D7 ROAS 20-40%: Healthy for most mobile games
- D30 ROAS 50-80%: On track for profitability
- D90 ROAS 100%+: Profitable acquisition
Factors Affecting ROAS
- CPI: Lower acquisition costs directly improve ROAS
- User Quality: Higher-quality users spend more
- Monetization: Better IAP/ad monetization increases revenue
- Retention: Users who stay longer generate more revenue
Strategies to Improve ROAS
- Focus on payer optimization: Optimize for purchase events, not just
installs
- Segment by value: Bid higher for segments with historically better ROAS
- Improve onboarding: Better Day 1 experience leads to more engaged (and
paying) users
- Optimize monetization: A/B test IAP pricing and ad placements
- Use predictive LTV: Predict user value early and optimize accordingly
Pro Tip: Don't compare ROAS across different apps or categories. A casual
game might have 30% D7 ROAS and be highly profitable, while a subscription app might need
100%+ D7 ROAS to break even. Know your unit economics.
← Back to article list
A/B Testing Creative: A Data-Driven Approach
9 min read • Updated Dec 2023
→
A/B Testing Creatives: A Data-Driven Approach
By Cubemobi Creative Team • 9 min read • Updated December 2023
Creative testing is essential for sustainable user acquisition. Without systematic testing,
you're leaving performance on the table. Here's how to run effective creative A/B tests.
Why A/B Test Creatives?
- Identify what resonates with your target audience
- Improve CTR, CVR, and ultimately CPI
- Combat creative fatigue proactively
- Build a library of proven creative concepts
What to Test
Focus on elements that can significantly impact performance:
- Concept: Gameplay vs. story vs. character-focused
- Opening Hook: The first 2 seconds are critical for video
- Call-to-Action: "Download Now" vs. "Play Free" vs. "Get Started"
- Art Style: Different visual treatments of the same concept
- Length: 15s vs. 30s vs. 45s video ads
- End Card: Static vs. animated end cards
Testing Framework
- Hypothesis: Define what you're testing and why
- Control: Use your best-performing creative as the baseline
- Variants: Create 2-4 variations changing one variable
- Sample Size: Run until you have statistically significant data
- Analysis: Compare results and draw conclusions
- Iterate: Build on winners, discard losers
Statistical Significance
Don't make decisions based on small sample sizes:
- Aim for at least 1,000 clicks per creative variant
- Run tests for at least 7 days to account for day-of-week variance
- Use statistical significance calculators to validate results
- 95% confidence level is the standard threshold
Common Mistake: Testing too many variables at once. If you change the
concept, colors, and CTA simultaneously, you won't know which change drove the result. Test
one variable at a time for clear learnings.
← Back to article list