Scaling TikTok ad spend for mobile apps is one of the trickiest challenges in mobile UA today. Move too fast and you reset the algorithm's learning phase, torching your CPIs overnight. Move too slowly and you miss your window before creative fatigue sets in.
In our experience, scaling TikTok campaigns for app installs from early-stage budgets to significant daily spend while maintaining stable cost per acquisition is achievable—but only with the right framework.
This guide walks you through the exact framework we use: how to increase budgets without breaking performance, when to duplicate ad groups versus scale existing ones, and how to layer in fresh creatives so the algorithm always has fuel.
Whether you're spending $200/day or $5,000/day, these steps will help you scale predictably.
Prerequisites: Before scaling, you need a baseline of proven performance. Specifically: at least one ad group that has exited TikTok's learning phase (typically 50 conversions in 7 days), a clear understanding of your target CPI or CPA, and a pipeline of creative variations ready to deploy.
You should also have your MMP properly configured with postback events (AppsFlyer, Adjust, etc.) so you can evaluate downstream quality, not just installs.
Page Contents
Step 1: Establish Your Performance Baseline Before Touching Budgets
Before you scale a single dollar, you need to know exactly what 'good' looks like. Document your current CPI, CPM, CTR, CVR, and most importantly, your cost per downstream event (subscription, purchase, etc.) at the ad group level. We typically need 7 to 14 days of stable data before we consider a campaign ready to scale.
Calculate your blended CPA across all ad groups
Don't just look at your best ad group in isolation. Your blended CPA across the campaign is the number you need to hold as you scale. Based on our experience managing post-ATT campaigns, blended channel-level CPAs are more reliable than campaign-level data for evaluating true performance.
Identify your winning ad groups and creatives
Tag every ad group as 'winner' (below target CPA with 50+ conversions), 'potential' (close to target, still learning), or 'loser' (above target with sufficient data). You'll only scale winners and tested potentials.
Set your daily conversion floor
TikTok's algorithm needs roughly 50 conversions per week per ad group to maintain optimization. If you're optimizing for installs, ensure each ad group you plan to scale is hitting at least 7 to 10 installs daily. For AEO campaigns optimizing on deeper events, consolidate until you're hitting at least 128 installs daily at the campaign level.
We've seen teams try to scale ad groups with only 20 conversions over a week. The algorithm hasn't learned enough at that point, and pumping budget into it just inflates CPIs. Patience here saves you thousands downstream.
Step 2: Scale Budgets Gradually Using the 20-30% Rule
The single most important rule in scaling TikTok ads: never increase an ad group's budget by more than 20 to 30% in a single day. TikTok's algorithm is extremely sensitive to budget changes. Large jumps (50%+) reset the learning phase, which means the algorithm re-enters exploration mode and your CPIs spike unpredictably.
Increase budgets once every 24 to 48 hours
If your ad group is spending $200/day profitably, increase to $240 to $260. Wait 24 to 48 hours for performance to stabilize. Then increase again. This compounds faster than you think: a 25% daily increase turns $200 into $600+ within a week.
Time your increases correctly
Make budget changes during low-traffic hours (typically early morning in your target geo). This gives the algorithm time to adjust pacing before the high-competition afternoon and evening auction windows.
Monitor for 'CPI creep' after each increase
After every budget bump, watch your CPI for the next 24 hours. A 10 to 15% CPI increase is normal and usually corrects. A 30%+ spike means you've pushed too hard. Roll back the budget and wait 48 hours.
In our experience, scaling in 20% increments on a 48-hour cadence tends to yield the most stable results on TikTok. It’s slower than Meta’s tolerance for budget changes, but TikTok’s auction dynamics are different—Meta’s higher daily budget increase limits.
Step 3: Duplicate Winning Ad Groups to Scale Horizontally
Vertical scaling (increasing budget on existing ad groups) has a ceiling. Eventually, CPIs start climbing because the ad group is saturating its audience segment. Horizontal scaling, duplicating your winning ad groups, lets you access fresh auction inventory without destabilizing what's already working.
Duplicate the ad group with identical settings
Create an exact copy of your winning ad group: same targeting, same creatives, same bid strategy, same budget. TikTok will enter a slightly different auction pool, and you'll often see CPIs match or beat the original for the first few days.
Stagger your duplicates
Don't launch 5 duplicates at once. Start with one duplicate, let it exit learning phase, then add another. Running too many identical ad groups simultaneously causes internal competition that drives up your own CPIs.
Vary one element per duplicate if needed
If you notice duplicates cannibalizing each other, introduce slight variations: swap one creative, adjust the age range by a few years, or shift the optimization event. This gives TikTok enough differentiation to find distinct user pockets.
We’ve seen horizontal scaling work particularly well when combined with broad targeting. As Matej Lancaric from SuperScale has discussed, broad targeting with large audiences because the algorithm has maximum room to find cheap conversions, and TikTok’s content-first algorithm handles this even better than interest-based approaches.
Step 4: Expand Targeting Strategically to Unlock New Audiences
Once you've scaled budgets and duplicated ad groups, the next lever is targeting expansion. But expanding targeting incorrectly is one of the fastest ways to waste budget. The key is controlled expansion: one variable at a time.
Start with broad and let TikTok's algorithm do the targeting
TikTok's algorithm, like Meta's, performs best with minimal targeting restrictions. If you're still running narrow interest-based targeting, test a fully broad ad group (age and geo only) with your best creatives. In our experience, broad often outperforms interest targeting on TikTok after the learning phase completes.
Expand to new geos methodically
If you're profitable in the US, test Canada, UK, or Australia in separate campaigns. Don't bundle new geos into existing campaigns because it disrupts the algorithm's learned behavior. Each new geo gets its own campaign, its own learning budget, and at least 2 weeks to prove itself.
Test lookalike audiences based on purchasers, not installers
If you use custom audiences, build lookalikes from your highest-value users (subscribers, purchasers), not all installers. Installer-based lookalikes cast too wide a net and often perform no better than broad targeting.
Remember that past purchase behavior is a far stronger performance predictor than contextual signals. This is why self-attributing networks like TikTok, which have rich behavioral data, tend to outperform context-based DSPs. Lean into the algorithm's strength by feeding it conversion data, not constraining it with narrow targeting.
Step 5: Feed the Algorithm with Fresh Creatives on a Weekly Cadence
Creative fatigue is the number one scaling killer on TikTok. Winning creatives can lose efficiency at higher spend levels as fatigue sets in over time. You need a steady pipeline of new variations, not just one or two winners you ride into the ground.
Maintain 3 to 5 active creatives per ad group
TikTok recommends 3 to 5 creatives per ad group, and in our experience this range works well in practice. Fewer than 3 leads to rapid fatigue. More than 5 fragments the algorithm's learning across too many assets. For a deeper breakdown, check out our guide on optimal number of creatives per group.
Introduce new creatives into existing winning ad groups
When you add a fresh creative to a proven ad group, TikTok will test it against your existing winners. This is safer than launching entirely new ad groups because the ad group already has learned optimization data.
Use RocketShip HQ's 4-Layer Hook System for new variations
Every new video creative should stack four hook layers in the first second: a visual pattern break (0.3 to 0.8s zoom or movement), a text overlay under 15 words that creates a curiosity gap, a verbal/voiceover line that builds human connection, and audio or music that amplifies emotion. This combination overrides the user's scrolling behavior and dramatically lifts thumbstop rates.
Rotate concepts, not just edits
Changing the background color or text font is not a new creative. You need genuinely different angles: new hooks, new value propositions, new formats (UGC vs. motion graphics vs. screen recordings). Test a new concept every week while iterating on winning concepts with minor variations—remember that hook determines script performance.
TikTok's creative lifespan at scale is significantly shorter than Meta's. At $5,000+/day, plan to introduce 3 to 5 net-new creative concepts per week. This sounds aggressive, but it's the cost of scaling on TikTok. The ideal ad length for app installs tends to be shorter than most teams expect, so production volume is achievable.
Step 6: Monitor and React to Learning Phase Resets
Every time TikTok's learning phase resets on an ad group, you lose 2 to 5 days of optimized delivery. Understanding what triggers resets, and avoiding them, is critical to sustained scaling. The most common triggers are budget increases over 50%, bid changes, targeting edits, and creative additions (in some cases).
Track learning phase status daily
In TikTok Ads Manager, check the 'Delivery' column for each ad group. If it shows 'Learning' after a change you made, note the date and monitor for 3 to 5 days before making additional changes.
Avoid stacking changes
Never change budget AND targeting AND creatives at the same time. Make one change, wait for the ad group to stabilize (24 to 48 hours minimum), then make the next change. Stacking changes makes it impossible to diagnose what caused a performance shift—and learning phase volatility until the algorithm re-stabilizes.
Know when to kill and restart
If an ad group has been in learning phase for over 7 days without exiting, it's unlikely to recover. Pause it, duplicate with fresh settings, and start again. Don't waste budget nursing a broken ad group back to health.
We've learned the hard way that even 'small' changes like updating a CTA button can reset learning on TikTok. When in doubt, duplicate the ad group with the change rather than editing the live one.
Step 7: Diversify to Other Channels Only After TikTok Is Saturated
The temptation to spread budget across 4 to 5 channels early is strong but usually counterproductive. Each ad channel needs sufficient conversion volume for its algorithm to learn. Spreading $1,000/day across TikTok, Meta, Snap, Google, and Apple Search means no channel gets enough data to optimize properly.
Max out TikTok first
As a general rule, early-stage apps should concentrate ad budgets on one or two self-attributing networks rather than spreading thin. If TikTok is your best channel, scale it until CPIs start rising consistently despite fresh creatives. That's your signal to diversify.
Add one channel at a time with dedicated budget
When you do diversify, give the new channel (Snap, Google UAC, or Apple Search Ads) its own budget allocation and at least 2 to 4 weeks of learning time. Don't pull budget from a working TikTok campaign to fund a test on another platform.
In practice, teams running lower daily spends often see better results concentrating on one or two channels rather than spreading budget thin across multiple channels.
Common Mistakes to Avoid
- Increasing budgets by 50%+ in a single day: This is the most common scaling mistake we see. A large budget jump resets TikTok's learning phase, causing significant CPI spikes. Stick to 20 to 30% increases max, with 24 to 48 hours between adjustments.
- Launching too many duplicate ad groups simultaneously: Running 5+ identical ad groups causes you to bid against yourself in TikTok's auction. Your CPIs rise because you're competing with your own campaigns. Stagger duplicates, launching one every 3 to 5 days.
- Neglecting creative refresh at higher spend levels: A creative that works at a lower daily budget will fatigue significantly faster at higher spend levels because it's shown to a much larger audience in the same timeframe. Plan for 3 to 5 new concepts per week at scale.
- Editing live winning ad groups instead of duplicating: Changing targeting, bids, or optimization events on a performing ad group risks resetting its learning phase. Always duplicate the ad group, make your change on the duplicate, and pause the original only if the duplicate proves out.
- Optimizing for installs when you should be optimizing for deeper events: As you scale, install-optimized campaigns can attract low-quality users who never convert downstream. Once you have sufficient volume (128+ installs/day), test AEO campaigns optimizing for purchase or subscription events. The CPIs will be higher, but the ROI usually improves significantly.
Scaling TikTok ad spend for mobile apps comes down to disciplined patience: gradual budget increases, strategic horizontal duplication, continuous creative production, and resisting the urge to make too many changes at once. Teams that scale budgets without scaling creative velocity typically hit a wall at higher spend levels because creative fatigue catches up.
Start by proving performance at a smaller budget, then scale vertically with the 20 to 30% rule, horizontally with duplicated ad groups, and creatively with fresh concepts every week using frameworks like the 4-Layer Hook System.
Once TikTok hits a natural ceiling, that's when you layer in additional channels with concentrated, dedicated budgets. This exact playbook can help app advertisers scale daily budgets on TikTok while working toward target CPAs. The key is treating scaling as a process, not an event.
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
- TikTok Ads for app growth: the complete guide (comprehensive guide)
- Best call-to-action buttons for TikTok
- How Many Creatives Do You Need Per TikTok Ad Group?
- What Is the Ideal TikTok Ad Length for App Installs?
- TikTok Ads for app growth: the complete guide
