Media planning for mobile user acquisition means deciding where the next dollar goes across channels based on post-install performance, not just cheap installs. The core job is to define roles for each channel, test into them with clean measurement, then move budget toward where you see reliable retention, CPA and ROAS.
If you treat your plan as a flexible portfolio instead of a static split, you can shift spend weekly as cohorts mature, channels saturate, or attribution rules change. The aim is a mix of performance, learning, and resilience so a policy change or platform wobble does not kill your funnel.
What is a practical structure for media planning for mobile user acquisition?
Start with a portfolio view. You are not buying isolated campaigns; you are building a system where each channel has a job:
- Prospecting at scale (programmatic, social, UGC) to find new audiences
- High-intent inventory (rewarded, search, OEM) to capture demand
- Upper-funnel reach (CTV, some display) to seed future converting traffic
For most apps, a practical starting split for net-new UA (excluding brand and retargeting) looks like this:
- 40–60% programmatic / in-app performance inventory
- 20–40% self-attributing networks (Meta, Google App Campaigns, TikTok, etc.)
- 10–20% rewarded and offerwall
- 5–15% OEM and preloads in Android-heavy markets
- 5–15% CTV for apps with enough LTV and good attribution
These are starting points, not targets. Your plan should define ranges, not fixed shares, so you can move budget weekly as data comes in.
How should you define objectives before allocating budget?
If the objective is fuzzy, your media plan will drift. You need channel-level objectives and constraints before you spend.
Common primary objectives by app stage:
- Early launch: maximize qualified volume at break-even D30 ROAS or payback under 60–90 days
- Growth: hit a specific D7/D30 ROAS or CPA while scaling weekly installs by 10–30%
- Late scale: keep margin stable while defending market share and fighting churn
Translate that into measurable targets:
- ROAS targets: D1 for casual games, D7 for subscriptions, D30–D90 for midcore/hardcore games
- CPA / CAC caps tied to payback window
- Retention floors (e.g., D7 ≥ 15% for casual, ≥ 25% for utility, whatever is realistic for your genre)
Then assign objectives by channel, for example:
- Programmatic: scale at target D7 ROAS with strict fraud controls
- Rewarded: hit a tighter CPA target since users show higher install intent
- CTV: maintain prospecting CPIs that lead to acceptable blended ROAS across devices
- OEM: buy low-CPI volume but protect quality with retention and event-rate checks
Your IOs and internal briefs should say this explicitly so buyers and algorithms optimize to the same outcome.
How should different channels contribute to your UA goals?
Each channel behaves differently on cost, intent, and measurability. Design the media plan so channels complement each other instead of competing blindly.
Here is a simplified comparison.
| Channel type | Typical role in UA plan | Strengths | Watchouts |
|---|---|---|---|
| Programmatic mobile | Scalable prospecting | Broad reach, granular bidding, creatives | Needs strong fraud control, learning time |
| Self-attributing | Prospecting + intent | Strong algorithms, rich signals | Opaque data, SKAN constraints on iOS |
| Rewarded / offerwall | High-intent conversions | High CVR, predictable event volume | Volume can be limited, incentive-driven use |
| OEM / preload | Always-on volume in Android | Low CPI, presence on device | Mixed quality, slower payoff |
| CTV | Upper/mid-funnel for installs | Big-screen impact, incremental reach | Requires solid cross-device attribution |
When you plan, write down the job of each line item:
- "Programmatic: test 3 exchanges for scale at target D7 ROAS."
- "CTV: reach new audiences and drive measurable installs at target blended CAC."
- "Rewarded: stabilize event volume for optimization and in-app experiments."
If a channel does not have a clear job, you will struggle to defend its budget.
How do you build test budgets into a media plan?
Most UA teams underfund tests. They spread tiny amounts across many experiments, then call everything inconclusive. A better pattern is to ringfence a real test budget inside the total spend.
A simple rule that works for many apps:
- 70–80% of budget in proven channels and setups
- 20–30% reserved for tests (new channels, new countries, new creative concepts, new bidding strategies)
For each test, define:
- Hypothesis: "CTV in US can hit D30 ROAS within 10% of Meta while adding incremental reach."
- Minimum spend: enough for meaningful data. For most SKAN or event-optimized campaigns, assume at least 50–100 post-install events or a few hundred installs per tactic before you judge.
- Duration: at least 7–14 days for short funnels, longer when your payback window is long.
Put those numbers into the media plan so test budgets do not get raided when one channel spikes in CPI for a day.
When should you increase, hold, or cut channel budgets?
Decisions to scale or cut should lean on cohort performance, not last-click CPI. A practical way to structure this is a weekly review using standardized metrics:
- D1 and D3 retention
- Early funnel event rate (e.g., registration, tutorial complete, level 3 reached)
- D7 ROAS or cost per key event
- Projected D30/D90 ROAS based on historical curves
Then apply simple rules by channel:
Increase budget when:
- Cohorts meet or beat target D7/D30 ROAS and are stable over at least 2–3 weeks
- CPIs are rising but ROAS is flat or up (quality is holding)
- You still see headroom: impression share low, frequency reasonable, creative fatigue limited
Hold budget when:
- Performance is near target but volatile, and you need another week of data
- You are mid-test and cohorts have not matured to the decision point yet
Reduce or pause when:
- D7/D30 ROAS is consistently below target and not improving with creative or targeting changes
- Incrementality tests show heavy cannibalization of organic or other paid channels
- Fraud indicators spike (suspicious CTIT patterns, abnormal retention) and cannot be cleaned with blocking
Use percentage moves, not all-or-nothing swings. For example, move budgets in 20–30% steps unless there is a clear measurement or fraud issue.
How do attribution differences affect media planning for mobile user acquisition?
Attribution differences across channels can distort performance views. Your media plan has to anticipate that.
Key issues:
- Self-attributing networks (SANs) vs MMP: SANs often over-claim against your MMP. Plan and report using one source of truth, usually the MMP, then monitor SAN dashboards for optimization only.
- SKAdNetwork / AdAttributionKit on iOS: delayed, aggregated postbacks and limitations on campaign counts mean you need longer test windows and fewer, cleaner setups. AppsFlyer and Adjust both emphasize keeping SKAN structures simple to avoid fragmentation.
- CTV and web-to-app: you often track impact with probabilistic models or household/device graphs from an MMP like Branch or AppsFlyer. Effects appear as lifts in blended metrics, not 1:1 last-click chains.
Media planning decisions should therefore use tiers of evidence:
- Direct, deterministic: click-through installs and events within attribution windows
- Modeled: SKAN cohorts, CTV probabilistic attribution, MMM-type outputs if you have them
- Blended: overall CAC/ROAS at portfolio level
For channels with weaker attribution (CTV, some programmatic web), set expectations differently. You might accept a looser ROAS target but require a clear lift in blended installs and revenue when the channel is on versus off.
How long should you wait for results before adjusting the plan?
Many teams move budget too fast. The right waiting time depends on your funnel length and attribution.
Simple guidelines:
- Apps with fast payback (casual games, utilities): evaluate channel changes after 7–14 days, since most value appears by D7–D14.
- Subscription and productivity apps: wait until D14–D30 cohorts before big budget reallocations, especially when trials delay revenue.
- Midcore/hardcore games and complex fintech: you may need to project D60–D90 performance using early KPIs and historical curves, then validate once full cohorts mature.
Document this in the media plan as a "decision cadence" per channel. For example:
- Meta, TikTok, Google App Campaigns: weekly adjustments based on D3–D7 event and ROAS data.
- Programmatic with SKAN: every 2 weeks, because of delayed postbacks and signal sparsity.
- CTV: monthly, using blended metrics and incremental tests.
That way, no one expects a CTV test to be judged on D3 CPI, and no one leaves an obviously unprofitable programmatic source running for 2 months.
How do you keep the plan flexible while avoiding chaos?
A flexible plan is not random. You need guardrails so the budget can move without wrecking learnings.
Practical guardrails:
- Budget bands per channel: e.g., programmatic 40–70% of total, SANs 20–40%, tests 10–30%. Move inside bands before rewriting them.
- Minimum viable budgets: never run a campaign or channel below the threshold where the algorithm can learn. If that is £500–£1,000 per week for a tactic, write it into the plan.
- Change limits per week: for mature channels, do not change budgets more than 50% week over week unless there is a data or integrity issue.
- Clear ownership: someone owns each channel's P&L and knows when and why budgets move.
You end up with a living media plan document that:
- States objectives and targets per channel
- Defines budgets and ranges
- Lists active tests and decision dates
- Logs each major budget move with a reason
Treat that document as the single point of truth for UA, product, and finance conversations.
FAQ
What is media planning for mobile user acquisition in simple terms?
Media planning for mobile user acquisition is the process of deciding how much budget to spend across channels like programmatic, social, CTV, rewarded and OEM to acquire users at target CPA or ROAS.
How much budget should I allocate to testing new UA channels?
A practical range for most apps is 20–30% of total UA budget reserved for structured tests across new channels, markets, bidding strategies or creative concepts.
When should I shift budget away from a channel with cheap CPI?
Shift budget away when cheap CPI does not translate into acceptable retention, CPA or ROAS. If D7/D30 cohorts are below target after basic optimization and fraud checks, low CPI is not helping your business.
How does CTV fit into a mobile UA media plan?
CTV usually acts as an upper or mid-funnel channel that drives incremental installs and revenue measured through cross-device attribution and blended metrics, rather than pure last-click performance.
How often should I update my UA media plan?
Most teams review and update their media plan weekly for short-funnel channels and monthly for slower-maturing or upper-funnel channels like CTV, adjusting budgets based on cohort performance and test results.