Should you use tcpi, tcpa, or troas bidding in Google App Campaigns? (2026)

27 min read | Updated April 2026

Choosing between tCPI, tCPA, and tROAS bidding in Google App Campaigns is the single highest-leverage decision in your campaign setup. Get it wrong and you burn budget on users who never convert. Get it right and Google's algorithm works for you, pulling in exactly the users your business needs.

This guide breaks down when each bid strategy wins, the conversion thresholds that make or break performance, and the real-world trade-offs practitioners face in 2026.

What are the three Google App Campaign bid strategies and how do they differ?

Google App Campaigns offer three bid strategies: tCPI (target cost per install), tCPA (target cost per action), and tROAS (target return on ad spend). Each optimizes for a different depth in the user funnel, from raw installs to revenue-weighted conversions.

Key insight: tCPI buys volume, tCPA buys actions, tROAS buys profitable revenue.

• tCPI: optimizes for install volume only
• tCPA: optimizes for a specific post-install event
• tROAS: optimizes for revenue-weighted user value
• Each requires progressively more conversion data to function

tCPI tells Google "get me installs at roughly this price." The algorithm optimizes purely for download volume without considering what happens after install. According to Google's official App campaign documentation, tCPI campaigns focus on maximizing installs.

tCPA shifts optimization deeper into the funnel. You define a post-install event (subscription start, purchase, level 10 completion) and Google targets users likely to complete that action.

Per Google App Campaign targeting mechanics, the algorithm uses signals like search history, app usage patterns, and device data to predict which users will convert.

tROAS is the most sophisticated strategy. Instead of treating all conversions equally, it weights users by their predicted revenue value. A user who subscribes to an annual plan at $99.99 gets prioritized over a monthly subscriber at $9.99.

According to AppsFlyer's State of App Marketing report, tROAS campaigns typically show 15-30% higher LTV per acquired user compared to tCPA campaigns in the same app category.

When should you use tCPI bidding in Google App Campaigns?

Use tCPI when you need raw install volume at scale, particularly during launch phases, seasonal pushes, or when you lack sufficient post-install conversion data for tCPA or tROAS.

According to Adjust's State of App Growth report, apps in the first 90 days post-launch rely on tCPI for 60-70% of their Google spend.

Key insight: tCPI is your launch weapon, not your long-term strategy.

• Launch phases with no conversion history
• Ad-monetized apps where installs equal revenue
• Chart-ranking burst campaigns
• Bootstrapping data for future tCPA migration
• Seasonal pushes needing maximum reach

The primary advantage of tCPI is simplicity and speed. Google's algorithm needs very little learning data to optimize installs. You can launch a tCPI campaign and see meaningful volume within 24-48 hours, compared to the 7-14 day learning period that tCPA and tROAS campaigns often require.

The core weakness is obvious: installs are not revenue. A tCPI campaign optimized at $1.50 might deliver thousands of installs where only 2-3% of users ever reach a monetization event.

Per data.ai's 2025 benchmarks, the average install-to-purchase rate across non-gaming apps sits at roughly 3.2%, but tCPI campaigns often skew below this because the algorithm has no incentive to find buyers.

There are three scenarios where tCPI remains the right call. First, new apps with zero post-install event data. You cannot run tCPA without conversion history, so tCPI bootstraps your measurement infrastructure. Second, apps monetized primarily through ads (not IAP or subscriptions) where install volume directly correlates with revenue.

Third, short-term burst campaigns for chart ranking or promotional events where you need maximum installs in a compressed window.

One pattern that works well: run tCPI for 2-4 weeks to accumulate conversion data, then migrate to tCPA once you have at least 10 in-app actions per day flowing through your measurement partner. This staged approach avoids the cold-start problem that kills tCPA campaigns launched prematurely.

What is a good tCPI benchmark for Google App Campaigns?

According to Liftoff's 2025 Mobile Ad Creative Index, median tCPI on Google App Campaigns sits at $1.22 for Android globally and $2.54 for iOS. Gaming apps trend lower at $0.75-$1.40 on Android, while finance and subscription apps often exceed $3.00 per install.

These benchmarks shift significantly by geography. Southeast Asian markets deliver Android installs at $0.15-$0.40, while US and UK installs run $2.00-$4.50 depending on category. Your tCPI target should be reverse-engineered from your blended LTV, not set based on industry averages.

When is tCPA the right bid strategy for Google App Campaigns?

tCPA is the right choice when you have a clear post-install conversion event, sufficient historical data (10+ conversions per day per Google's recommendation), and your app monetizes through subscriptions or in-app purchases where not all installs are equal.

According to industry analysis on subscription app acquisition, trial-to-conversion drop-off averages 50-60%, making tCPA essential for filtering out users who will never pay.

Key insight: tCPA campaigns need 10+ daily conversions to exit learning mode reliably.

• Subscription apps optimizing for trial or purchase
• Apps with 10+ daily post-install conversions
• When install quality matters more than volume
• IAP-heavy gaming apps targeting first purchase
• Set tCPA within 10-15% of observed CPA for stability

The magic of tCPA is that Google's algorithm learns which user profiles actually complete your target action, then bids aggressively for lookalikes. For subscription apps, this means setting your conversion event to "start trial" or "subscribe" rather than "install." The algorithm then ignores cheap-install users who bounce at the paywall.

The conversion threshold is non-negotiable. Google's own documentation states 10 conversions per day as the minimum for stable optimization. In practice, campaigns with fewer than 30-50 weekly conversions oscillate wildly, with CPA swings of 40-60% day over day. If your app generates only 5 purchases per day, tCPA will struggle.

Consider using a higher-funnel event (like "add to cart" or "start trial") as your optimization target.

One critical nuance for subscription app optimization: the event you optimize for dramatically shapes your user quality. Optimizing for "trial start" gets volume but may attract users who cancel before conversion. Optimizing for "paid subscription" gets higher-quality users but requires more data and higher CPAs.

Most subscription apps find the sweet spot optimizing for trial starts while monitoring the trial-to-paid ratio as a guardrail metric.

The tCPA you set should reflect your true willingness to pay for a conversion, not an aspirational number. Setting tCPA 20%+ below your actual break-even CPA starves the algorithm of budget and collapses volume.

Per AppsFlyer's Performance Index analysis, campaigns with tCPA set within 10-15% of the actual observed CPA deliver the most stable scale.

What happens if you don't meet the minimum conversion threshold for tCPA?

The campaign enters a perpetual learning state. Google labels this "Learning (limited)" in the interface, and performance becomes erratic. According to Google's bidding documentation, campaigns stuck in learning limited for more than 14 days should be restructured.

The fix is straightforward: move to a higher-funnel event. If you get 4 purchases per day but 25 trial starts, optimize for trials. Alternatively, consolidate ad groups. Splitting campaigns into too many geo or creative segments dilutes conversion volume. Each campaign needs its own 10/day threshold.

When should you switch to tROAS bidding in Google App Campaigns?

Switch to tROAS when your users have meaningfully different revenue values and you can pass accurate revenue data back to Google. According to AppsFlyer's eCommerce App Marketing report, apps using tROAS bidding see 18-25% higher revenue per user compared to flat tCPA campaigns in the same vertical.

Key insight: tROAS only outperforms tCPA when your user values genuinely vary.

• Users have meaningfully different revenue values
• Revenue data flows accurately via MMP/server-to-server
• 10+ daily conversions with value attached
• Ecommerce, tiered subscriptions, or whale-driven gaming
• Uniform monetization apps should stick with tCPA

tROAS tells Google "for every dollar I spend on ads, I want X dollars back in revenue." A 200% tROAS target means you want $2 in revenue for every $1 in ad spend. The algorithm then bids more for users predicted to generate high revenue and less for low-value users.

This strategy shines in three scenarios. Ecommerce apps where basket sizes range from $15 to $500+. Subscription apps with multiple tiers (monthly at $9.99 vs. annual at $99.99). Gaming apps with whale dynamics where 2-3% of users generate 50%+ of IAP revenue.

If your monetization is uniform (every user pays roughly the same), tROAS adds complexity without benefit.

The data requirements are steeper than tCPA. You need 10+ conversion events per day with revenue values attached. Per growth strategy analysis from Phil Carter, the "value capture" loop depends on accurately measuring which users generate the most revenue and feeding that signal back to acquisition channels.

Revenue data quality is the make-or-break factor. If your MMP (AppsFlyer, Adjust, etc.) sends inaccurate or delayed revenue events, tROAS optimization will misfire. Server-to-server revenue reporting with sub-24-hour latency is the minimum standard.

Apps using RevenueCat's subscription tracking or similar server-side infrastructure tend to see cleaner tROAS performance because renewal and refund data flows accurately.

How do you calculate the right tROAS target?

Start with your Day 7 or Day 30 ROAS benchmark from existing campaigns, then set tROAS 5-10% below that observed value. This gives Google room to optimize without being over-constrained.

According to industry data from Adjust's global benchmarks, subscription apps targeting Day 30 ROAS typically set initial tROAS targets between 80-150% depending on payback window.

Never set tROAS at your break-even point. The algorithm needs headroom. If your break-even ROAS is 120%, start with a tROAS of 100-110% and tighten gradually as the campaign matures and accumulates conversion data.

What are the minimum conversion thresholds for each Google App Campaign bid strategy?

Google recommends 10 conversions per day for tCPA and tROAS campaigns, with no minimum for tCPI. In practice, campaigns need 30-50 weekly conversions minimum for stable optimization, and 100+ weekly for tROAS to work reliably with value differentiation.

Key insight: The 10/day minimum is Google's floor, not the target for good performance.

• tCPI: no minimum conversion threshold
• tCPA: 10/day minimum, 30+/day for stability
• tROAS: 10/day minimum, 50+/day recommended
• Consolidate campaigns to concentrate conversion volume
• "Learning limited" after 14 days means restructure needed

Need help scaling your mobile app growth? Talk to RocketShip HQ about how we apply these strategies for apps spending $50K+/month on UA.

Google's official threshold of 10 conversions per day per campaign is documented in their App campaign help center. But this is the absolute minimum. Think of it as the point where the algorithm can technically function, not where it functions well.

At 10/day, expect significant CPA volatility. The algorithm is working with a tiny sample size and will make aggressive bets that sometimes pay off and sometimes blow up. At 25-30/day, performance stabilizes meaningfully. At 50+/day, you get the kind of consistent, predictable performance that lets you scale confidently.

For tROAS specifically, volume matters even more because the algorithm needs enough data to differentiate high-value from low-value users. With only 10 conversions per day, there may be just 1-2 high-value events in the mix, which is insufficient for pattern recognition.

Industry practitioners report that tROAS campaigns need roughly 2-3x the conversion volume of equivalent tCPA campaigns to achieve similar stability.

This is where campaign structure becomes critical. If you split US campaigns into 4 ad groups with different creatives, each ad group needs to independently meet the threshold. Consolidation is almost always better for algorithmic performance.

Use creative asset variety within fewer campaigns rather than spinning up separate campaigns per creative theme.

How should you transition from tCPI to tCPA or tROAS?

Transition in stages: run tCPI for 2-4 weeks to build conversion history, then launch a parallel tCPA campaign while maintaining the tCPI campaign at reduced budget. Cut tCPI only after tCPA consistently delivers 10+ daily conversions for at least 7 days.

Key insight: Never hard-switch bid strategies; run parallel campaigns during transition.

• Run tCPI 2-4 weeks to build conversion data
• Launch tCPA in parallel at reduced budget
• Ramp tCPA over 7-14 days, not overnight
• Cut tCPI only after tCPA stabilizes at 10+/day
• Weight performance signals by spend level during transition

The biggest mistake in bid strategy migration is the abrupt cutover. Pausing a tCPI campaign and launching tCPA in its place resets the campaign's learning history. Google treats the new campaign as starting from zero, regardless of the account's historical data.

The proven approach is parallel operation. Keep the tCPI campaign running at 50-60% of its original budget while ramping the tCPA campaign from $50-100/day up to full budget over 7-14 days. This gives the tCPA campaign time to exit learning mode without a gap in acquisition volume.

Monitor three metrics during transition. First, the tCPA campaign's conversion volume: is it hitting 10+/day consistently? Second, cost per conversion: is the actual CPA within 20% of your target? Third, downstream quality: are tCPA-acquired users showing similar or better retention and monetization versus tCPI users?

The same logic applies when moving from tCPA to tROAS. RocketShip HQ's Weighted Anomaly Scoring framework helps here: weight performance changes by spend level, not just percentage swings. A 15% ROAS drop on $5,000/day spend is a much bigger signal than a 40% drop on $200/day.

This prevents premature panic during the tROAS learning phase when daily fluctuations are expected.

One nuance specific to 2026: Google's campaign types now include "Maximize conversions" auto-bidding, which can serve as an intermediate step between tCPI and tCPA if your conversion volumes are borderline.

How does bid strategy choice affect creative testing in Google App Campaigns?

Bid strategy directly determines how aggressively Google rotates your creative assets. tCPI campaigns distribute impressions more evenly, making them better for creative testing. tCPA and tROAS campaigns concentrate spend on proven asset combinations, which limits creative exploration.

According to Google App Campaign creative asset guidelines, each campaign should include 20+ creative assets across text, image, and video formats.

Key insight: tCPI spreads impressions across creatives; tCPA/tROAS concentrate on winners.

• tCPI tests creatives broadly, good for discovery
• tCPA narrows to winners within 3-5 days
• Run a dedicated tCPI creative testing campaign at 10-15% budget
• Video assets generate 30-40% lower CPI per Liftoff data
• tROAS creative testing cycles take 20-30% longer

Google App Campaigns use a machine learning system that automatically assembles ads from your uploaded assets. Under tCPI, the algorithm's primary constraint is install cost, so it tests more combinations to find volume. Under tCPA, it narrows quickly to the asset combinations that drive conversions, often within 3-5 days.

This creates a practical tension. You want to test new creatives, but tCPA campaigns give new assets very little runway before deprioritizing them. The workaround: run a dedicated creative testing campaign on tCPI at 10-15% of your total Google budget.

Use it purely to identify winning assets, then port those winners into your tCPA or tROAS scaling campaigns.

Video assets get disproportionate weight in Google App Campaigns. According to industry benchmarks from Liftoff, video creatives in App Campaigns generate 30-40% lower CPI compared to static image ads. For tCPA campaigns, vertical video (9:16) with strong hooks in the first 3 seconds drives the highest conversion rates.

Similar creative principles apply across platforms, as discussed in before-and-after ad formats and fail-ad creative strategies.

For tROAS campaigns, creative testing becomes even more constrained because the algorithm is optimizing for revenue prediction, not just clicks or installs. High-value user targeting narrows the audience, which means fewer impressions and slower creative learning cycles. Budget 20-30% more time for creative iteration in tROAS campaigns versus tCPA.

Which bid strategy works best for subscription apps on Google App Campaigns?

tCPA targeting trial starts is the workhorse strategy for most subscription apps, delivering the best balance of volume and quality. According to mobile growth practitioners like Marcus Burke, trial optimization campaigns typically see 50-60% trial-to-conversion drop-off, making trial start the highest-signal event that still meets volume thresholds.

Key insight: Optimize for trial starts first; graduate to purchase optimization only when volume allows.

• Start with tCPA on trial starts, not paid subscriptions
• Need 10+ daily trial events for stable optimization
• Graduate to purchase optimization only at 10+ daily purchases
• tROAS only works with tiered pricing and 50+ daily conversions
• Web-to-app funnels use different campaign types entirely

Subscription app acquisition on Google follows a predictable maturity curve. At launch, tCPI to build data. At 10+ daily trials, switch to tCPA on trial starts. At 10+ daily paid subscriptions, test tCPA on subscription purchase. At 50+ daily subscriptions with tiered pricing, test tROAS.

Most subscription apps never reach the tROAS threshold because their daily conversion volumes are too low for revenue-weighted optimization. This is fine. tCPA on trial starts remains effective well into scale.

The critical optimization lever isn't the bid strategy itself but the paywall design and trial mechanics that determine what percentage of trialists convert to paid.

A common mistake: optimizing for "subscribe" (paid) events too early. If you only get 5 paid subscriptions per day from Google, the algorithm lacks the data to differentiate user profiles. You'll get wildly inconsistent CPAs and the campaign may never exit learning mode.

Stick with trial starts until paid subscription volume consistently exceeds 10/day for at least two weeks.

Web-to-app funnels add another dimension. Some subscription apps bypass the app store entirely for acquisition, using landing pages that capture email or start web trials before directing users to download.

Per insights from Blinkist's native ad landing page approach, this model enables faster creative iteration and avoids app store commission on initial conversion. However, web-to-app campaigns typically run on Google's standard Search or Display networks rather than App Campaigns, so they use different bid strategies.

How does bid strategy choice differ between gaming and non-gaming apps?

Gaming apps can leverage tROAS much earlier than non-gaming apps because IAP revenue events generate higher daily volumes and more variance in user value. According to AppsFlyer's Performance Index, gaming apps running tROAS see 22% higher Day 30 ROAS compared to gaming apps on tCPA.

Key insight: Gaming's whale dynamics make tROAS the natural end-state; non-gaming apps often stay on tCPA.

• Gaming: top 5% of payers drive 60-65% of IAP revenue
• tROAS adds 20%+ ROAS lift for gaming, 8-12% for non-gaming
• Hyper-casual games should stay on tCPI
• Subscription apps often plateau at tCPA on trial starts
• Optimize for first IAP in gaming, not installs

The fundamental difference is revenue distribution. In gaming, per data.ai, the top 5% of paying users generate roughly 60-65% of total IAP revenue. This extreme variance means tROAS can dramatically improve campaign efficiency by finding these high-spend users, something tCPA (which treats a $1 and $100 purchaser identically) cannot do.

Non-gaming apps, particularly subscription apps, have much flatter revenue distributions. A monthly subscriber and an annual subscriber differ by 5-10x in value, but the pool of annual subscribers is large enough that tCPA still captures them reasonably well.

The incremental benefit of tROAS is real but smaller, typically 8-12% higher revenue per user versus the 20%+ gaming apps see.

For gaming specifically, the optimal event progression looks different. Rather than trial starts, gaming apps typically optimize for "first IAP" or "Day 7 engagement" milestones.

According to cross-category advertising analysis, the right in-app event varies by monetization model: IAP-heavy games should optimize for purchase events, while ad-monetized games should stay on tCPI or use engagement events (like Day 3 retention) as tCPA targets.

Hyper-casual games represent a special case. Their average revenue per user is low ($0.05-$0.15 per install from ad monetization), making tROAS impractical because Google needs meaningful revenue variance to optimize effectively.

tCPI remains the dominant strategy for hyper-casual, with channel selection and creative velocity mattering more than bid strategy sophistication.

What are the most common tCPI, tCPA, and tROAS mistakes that waste budget?

The three costliest mistakes are setting targets too aggressively (starving the algorithm), splitting campaigns into too many segments (diluting conversion volume), and using the wrong optimization event (sending a weak signal).

Per Google's bid strategy troubleshooting guide, overly aggressive targets are the #1 reason campaigns stay in "Learning limited" status.

Key insight: Aggressive bid targets feel efficient but actually destroy volume and increase costs.

• Never set targets 20%+ below actual performance
• Consolidate campaigns to concentrate conversion data
• Optimize for deepest event meeting 10/day threshold
• Tighten bids 5% at a time, not in big jumps
• Avoid shallow events like app open or registration

Setting tCPA at $15 when your actual CPA runs $25 doesn't get you $15 CPAs. It gets you almost zero conversions. The algorithm can't find enough users at that price point and essentially stops spending. Then marketers panic, raise the bid, and the campaign re-enters learning mode.

This yo-yo pattern wastes 2-4 weeks of optimization history.

The fix is counterintuitive: start with a target 10-20% above your ideal CPA, let the algorithm find its rhythm, then tighten 5% every few days while monitoring volume. The same principle applies to tROAS. Set your initial target below your break-even ROAS to give the algorithm room.

Campaign fragmentation is the second killer. Marketers who create separate campaigns for US-Male-25-34 and US-Female-35-44 end up with 8 campaigns each getting 3 conversions per day instead of one campaign getting 24. Google App Campaigns already handle audience segmentation internally. Trust the algorithm with broad targeting and consolidated structure.

The third mistake is choosing the wrong optimization event. Some apps optimize for "app open" or "registration" because these events generate high volume, but they carry almost zero revenue signal. The algorithm dutifully finds users who open the app once and never return.

Choose the deepest funnel event where you can still meet the 10/day threshold. The balance between event depth and volume is the single most important configuration decision in any Google App Campaign.

Also common: ignoring the compliance requirements for certain verticals like fintech, which can lead to ad rejections that waste learning-phase spend.

How do Google App Campaign bid strategies compare to Meta and Apple Search Ads bidding?

Google's tCPA and tROAS map conceptually to Meta's cost cap and minimum ROAS bid strategies, but Google offers less manual control over audience targeting. Apple Search Ads uses a CPT (cost per tap) model that functions like tCPI with keyword-level control.

According to AppsFlyer's cross-network analysis, Google App Campaigns deliver 15-20% lower CPI but 10-15% lower Day 7 retention compared to Apple Search Ads.

Key insight: Google wins on volume and cost; Apple Search Ads wins on intent and quality.

• Google: algorithmic, broad reach, lower CPI
• Apple Search Ads: intent-based, higher CPI, better retention
• Meta: creative flexibility, audience control, cost cap bidding
• Google tCPA ≈ Meta cost cap in concept
• Diversify 40/30/30 across top channels

The fundamental difference is targeting philosophy. Google App Campaigns are almost entirely algorithmic. You upload creatives, set a bid strategy, and Google decides where, when, and to whom your ads appear across Search, Play Store, YouTube, Display, and Discover.

Meta gives you more audience control through interest targeting and custom audiences, though Advantage+ campaigns are moving toward Google's fully automated model.

Apple Search Ads occupies a different niche. Users are actively searching for apps, which means intent is inherently higher. CPT bids on Apple Search Ads average $1.50-$3.00 per tap according to Search Ads benchmark reports, with conversion rates from tap to install averaging 50-60%.

The resulting CPI is higher than Google, but downstream metrics (retention, trial rate, subscription rate) typically outperform.

For a mature app running multi-channel acquisition, the bid strategy framework looks like this. Apple Search Ads (CPT) captures high-intent demand. Google App Campaigns (tCPA/tROAS) drives efficient scale across Google's inventory. Meta (cost cap / min ROAS) provides creative testing flexibility and detailed audience insights.

Each platform's bid strategy exists in its own ecosystem and cannot be directly compared, but the underlying logic (pay for volume vs. pay for actions vs. pay for value) is universal.

Portfolio budgeting across these channels matters as much as individual bid strategy selection. As explored in best paid channels for mobile UA, allocating 40-50% of budget to the best-performing channel and 20-30% each to the other two provides both scale and diversification.

How should you monitor and adjust bid strategy performance over time?

Monitor on a 7-day rolling average, not daily snapshots. Adjust bids in 5-10% increments no more than once per week. According to Google's best practices, each bid change triggers a partial learning reset that takes 3-5 days to stabilize.

Key insight: Weekly monitoring with 5-10% bid adjustments beats daily micro-optimization every time.

• Monitor 7-day rolling averages, not daily snapshots
• Adjust bids 5-10% at a time, once per week max
• Each bid change triggers 3-5 day learning reset
• Never change bid target by more than 15% at once
• Q4 CPIs spike 15-25% due to seasonal demand

Daily performance in Google App Campaigns is inherently noisy. A campaign can show $12 CPA one day and $22 CPA the next, then average $16 over the week. Reacting to daily fluctuations by adjusting bids creates a destructive feedback loop: change bids, trigger learning, see noise, change bids again.

RocketShip HQ's Weighted Anomaly Scoring approach is useful here: weight metric changes by spend level. A 15% ROAS drop on a campaign spending $5,000/day is far more actionable than a 40% drop on a $200/day campaign.

This simple heuristic (abs(% change) × sqrt(spend)) eliminates over 70% of false alarms in performance monitoring.

The cadence that works: review dashboard daily for catastrophic issues (spend dramatically above or below target, campaign paused by policy violation). Make bid adjustments weekly based on 7-day rolling CPA or ROAS.

Review conversion event health monthly to ensure your MMP is reporting accurately and conversion volumes haven't drifted below thresholds.

When adjusting, follow the 15% rule: never change your bid target by more than 15% in a single adjustment. Larger changes force a full learning reset.

If you need to move from $20 tCPA to $30 tCPA, do it in three steps over three weeks: $20 → $23 → $26 → $30.

Also track macro-level shifts. According to Adjust's global data, CPIs increase 15-25% during Q4 (October-December) due to holiday advertising demand. Build seasonal bid adjustments into your plan rather than reacting after costs spike.

The bid strategy decision tree is straightforward once you know your conversion volume. Start with tCPI to build data, graduate to tCPA when you hit 10+ daily conversions, and test tROAS only when your user values genuinely vary and volume exceeds 50 daily events.

The most common failure mode is premature sophistication: jumping to tROAS before you have the data to support it. Start simple, monitor weekly, adjust in 5-10% increments, and let the algorithm learn.

Frequently Asked Questions

Can you run tCPI and tCPA campaigns simultaneously in the same Google Ads account?

Yes, and it's actually recommended during transitions. Running parallel tCPI and tCPA campaigns doesn't cause self-competition because Google's auction system deduplicates at the ad group level. Per Google's documentation, the highest-performing ad enters each auction. Maintain tCPI at 40-50% of original budget while scaling tCPA.

Does tROAS work for apps with only one subscription price point?

Poorly. tROAS requires revenue variance to add value. If every converting user generates exactly $9.99, the algorithm can't differentiate high-value from low-value users and tROAS degrades to tCPA behavior with extra complexity. Stick with tCPA unless you have at least 2-3 meaningfully different price tiers.

How long should you wait before evaluating a new bid strategy's performance?

Allow a minimum of 14 days and 100+ conversions before drawing conclusions. According to Google's campaign optimization guides, the learning phase typically lasts 7-10 days, and you need at least one full week of post-learning data to evaluate. Premature judgment leads to abandoning strategies that would have worked.

What happens to campaign performance when you change bid strategies mid-campaign?

Changing bid strategy within an existing campaign triggers a full learning reset, equivalent to starting a new campaign. Performance typically drops 20-40% during the 7-14 day re-learning period per Google's documentation. This is why creating a new campaign with the new bid strategy and running it in parallel is strongly preferred over modifying an existing campaign.

Should you use the same optimization event for both Google and Meta campaigns?

Not necessarily. Each platform's algorithm learns differently. Meta's trial optimization often skews toward younger demographics (18-24) as noted in subscription app advertising research, while Google's audience signals are different. Test the same event on both, but don't assume identical performance. Some apps optimize for trial on Meta and purchase on Google.

How does Google's machine learning handle tCPA for apps with long conversion windows?

Google's prediction models work best with conversion events that occur within 1-3 days of install. For events with 7+ day delays (like subscription renewal after a free trial), the algorithm struggles because the feedback loop is too slow. Per industry best practices, use a proxy event (trial start, registration) and monitor downstream conversion rates manually.

Do Google App Campaign bid strategies work differently on iOS versus Android?

Yes. iOS campaigns receive less user-level signal data due to Apple's App Tracking Transparency framework. According to AppsFlyer's cross-platform data, iOS tCPA campaigns show 25-35% higher CPA volatility compared to Android campaigns with identical settings due to reduced signal availability.

Is there a minimum daily budget required for each bid strategy?

Google recommends a daily budget of at least 50x your tCPI or 10x your tCPA target. For a $20 tCPA campaign, that means $200/day minimum. Campaigns below this threshold cannot generate enough daily conversions to exit the learning phase. For tROAS, plan $300-500/day minimum to feed the value-optimization model.

Looking to scale your mobile app growth with performance creative that delivers results? Talk to RocketShip HQ to learn how our frameworks can work for your app.

Not ready yet? Get strategies and tips from the leading edge of mobile growth in a generative AI world: subscribe to our newsletter.

Related Reading

• The complete guide to mobile user acquisition (comprehensive guide)
• Ad compliance for fintech app marketing (2026)
• Adjust State of App Growth Report: Global Trends and Benchmarks (2026)
• How to advertise a music or audio streaming app (2026)
• How to advertise a social networking app (2026)