Most mobile user acquisition advice still describes a world that ended with App Tracking Transparency. You cannot see most of your iOS conversions at user level, your best channel optimises on a signal you partly control, and the creative is doing the targeting that audiences used to do.
This guide covers how user acquisition actually works now: what to fix before you spend, which channels do what, why creative became the main lever, and how to bring acquisition costs down without simply buying cheaper and worse users.
What Is Mobile User Acquisition?
Mobile user acquisition (app user acquisition, or UA) is the paid part of getting people into your app. It sits alongside app store optimization, which converts the traffic once it arrives, and product work, which decides whether the user stays.
Two things changed and they changed everything downstream.
Measurement became probabilistic. After ATT, most iOS installs arrive without a user-level identifier. You work with SKAdNetwork postbacks, modelled conversions and aggregated reporting. Decisions that used to be made on exact numbers are now made on estimates with confidence intervals, and teams that have not adjusted their process to that are making confident decisions on noise.
Targeting moved into the creative. Broad targeting plus a strong creative now outperforms narrow audience targeting on every major platform, because the platforms model the audience better than you can specify it. The creative is what tells the algorithm who the ad is for.
Before you spend: three numbers that decide everything
Most failed UA programmes failed before the first campaign because nobody agreed what success was.
Cost per acquisition, measured at the right event. Cost per install is a vanity number for a subscription app. The number that matters is cost per trial, cost per paying user or cost per deposit, depending on your model. Optimising to installs reliably produces cheap installs and no revenue.
Payback period. How many months until a cohort returns what it cost to acquire. You can estimate it with our free payback period calculator. This, not ROAS on day one, is what determines how fast you can scale, because it sets how long your cash is tied up.
Lifetime value, by cohort and by channel. A blended LTV hides the fact that one channel brings users worth triple another. You need it split, and you need it early enough to act on, which usually means a day 7 or day 30 proxy rather than waiting a year for the real figure.
If you cannot state these three, the first task is not launching campaigns. It is measurement.
Measurement comes first
A mobile measurement partner, configured properly. An MMP attributes installs and in-app events across channels. The configuration is where the value is: if your events are not mapped to the moments that matter commercially, you have attribution for the wrong things.
A deliberate SKAdNetwork schema. You get a limited number of conversion values and a short measurement window. What you encode into them decides what you can optimise toward for the rest of the year. Most teams accept a default schema and then wonder why iOS optimisation is blunt.
Web-to-app, where it fits. Routing paid traffic through a web funnel before the store gives you first-party data, your own checkout and a measurable step that ATT does not obscure. It is not right for every app, and it adds a conversion step, but for subscription apps it often pays for itself. Our work on Workout for Women produced 27x subscription growth and a 5x reduction in cost per acquisition on web2app landings specifically.
User Acquisition Channels and What Each Is For
Channels are not interchangeable budget destinations, and user acquisition campaigns on each need a different setup. Each is good at a different job.
Apple Ads captures intent. Someone searching the App Store has already decided they want something in your category, which is why it usually carries the best conversion rate and the lowest cost per paying user. It is also capped by search volume, so it scales to a point and then stops. Custom product pages are the main lever: matching the store page to the keyword intent moves the numbers more than bidding does.
Meta creates demand. It is the volume channel for most consumer apps and the one most sensitive to creative quality and fatigue. Expect to replace winning creatives in weeks rather than months.
Google App Campaigns spans search, Play, YouTube and display from one campaign, which makes it efficient and opaque in equal measure. You feed it assets and conversion signals and it decides the rest, so asset quality and signal quality are the entire job.
TikTok rewards native, fast, creator-style content and punishes repurposed Meta assets. It can be very efficient for the right verticals and consistently poor for the wrong ones.
Programmatic and networks add incremental volume once the core channels are saturated, with more variance in quality and more need for fraud controls.
| Channel | What it is for | Cost position | What a meaningful test needs | How fast you can read it |
|---|---|---|---|---|
| Apple Ads | Captures intent that already exists in App Store search | Usually the lowest cost per paying user, capped by search volume | Brand terms plus a short category keyword list, and one custom product page per intent cluster | Days. Keyword-level reporting, with no modelled delay |
| Meta | Creates demand. The volume channel for most consumer apps | Mid, and it rises as creative fatigues | An optimisation event at trial or purchase, and four genuinely different creative concepts rather than variants | Creative verdict in one to two weeks. Value verdict only after a full payback cycle, and on iOS add the SKAdNetwork postback delay |
| Google App Campaigns | One campaign across Search, Play, YouTube and display | Opaque. Moves with asset and signal quality more than with bids | A complete asset set per group, and in-app events mapped to the moments that matter commercially | Slower. It reallocates internally and placement visibility is limited |
| TikTok | Native, creator-style demand. Strong in some verticals and consistently weak in others | Low cost per install, wide variance in value | Assets made for the platform rather than repurposed from Meta, and a vertical-fit check before budget | Creative verdict fast, quality verdict needs a payback cycle |
| Programmatic and networks | Incremental volume once the core channels are saturated | Lowest cost per install, highest variance in value | Source-level reporting and a working blocklist process | Slow and high variance. Judge by source, not by channel |
The sequencing that works for most subscription apps: prove the funnel on Apple Ads where intent is highest, scale on Meta once the paywall and creative hold up, add Google and TikTok for volume, then consider programmatic.
App store optimization is part of acquisition, not next to it
Every paid install passes through your store listing. If it converts at 25% and a competitor's converts at 40%, you pay a 60% premium on every install for the same spend and the same creative.
This is the most commonly ignored multiplier in UA, usually because ASO and paid sit with different people or different vendors. The two decisions that matter most: the first screenshot, which does most of the conversion work, and custom product pages matched to keyword or campaign intent.
Creative is the targeting lever
Since the platforms model audiences better than manual targeting can specify them, the creative became the instrument that tells them who to find.
Volume with a decision framework. Volume alone produces expensive noise. A framework without volume produces nothing to choose between. What works is a steady production rate plus a clear rule for what counts as a winner, measured past the install.
Concepts, not variations. Ten colour variants of one idea test nothing. Four genuinely different angles on why someone would use the app test something.
Measure creative on downstream quality. A creative with a high install rate and poor trial-to-paid conversion is actively harmful, because it teaches the algorithm to find more of the wrong people. On NEMO we produced and tested over 700 creatives, and the result that mattered was cost per deposit falling 8x on Meta, not the click-through rate.
How to reduce user acquisition costs
Lower CAC usually comes from conversion and measurement, not from cheaper media.
Know your benchmark before you judge your number. CAC is meaningless in isolation. It is meaningful against your payback target and against what your category pays. A high CAC with a two-month payback is a better business than a low CAC with no payback at all.

Fix conversion before you cut bids. Store listing, onboarding and paywall each multiply the cost of every install above them. A paywall improvement lowers effective CAC across every channel at once, which no bid change can do. On TouchRetouch a single custom product page aligned to the campaign's intent cut cost per trial by 32.6%.
Tighten the optimisation event. If you optimise to installs you will get installs. Moving the optimisation event down the funnel, to trial or purchase, raises the nominal cost per install and lowers the cost per paying user. Those are not the same direction and only one of them matters.
Prioritise the audiences that already pay. Cohort analysis usually shows a minority of segments producing most of the revenue. Concentrating spend there lowers blended CAC without touching a bid.
Use retargeting for the gap between install and payment. Many users install and never convert. Re-engaging them is almost always cheaper than acquiring a replacement.
Referral where the product supports it. A referral programme lowers blended CAC when sharing is natural to the product. When it is bolted onto a product nobody shares, it produces discounts and no growth.
Scaling without breaking the economics
Add a channel when the current one is saturated, not when it is merely working. Saturation looks like rising cost per paying user at flat spend, or flat volume at rising spend.
Scale in steps your measurement can keep up with. Doubling spend overnight gives you a cohort you cannot read for a fortnight, and by the time you can, the money is spent.
And watch payback rather than daily ROAS while scaling. Daily ROAS falls when you grow, which is expected. Payback breaking is the signal to stop.
Common mistakes
- Optimising to cost per install for a subscription app
- Accepting the default SKAdNetwork schema and never revisiting it
- Running paid traffic onto an unoptimised store listing
- Treating creative as a design task rather than a performance one
- Judging a channel before a full payback cycle has completed
- Adding channels for diversification before the first one is saturated
Conclusion
Mobile user acquisition in 2026 is three disciplines in a trench coat: measurement you can trust, creative that does the targeting, and conversion that makes every install worth more. The media buying is the easy part and the part that matters least.
Start by agreeing what you optimise toward and whether you can measure it. Everything else follows from that.
Paid channels are where most of this budget goes, and where most of it leaks. If you'd rather hand the buying to a team that scales on cost per paying user, talk to our mobile user acquisition agency. If the store page and paywall need work too, our app marketing agency runs all of it as one plan.
FAQ
What is a mobile user acquisition strategy?
It is the plan for how you pay to get users into your app and how you judge whether it worked. In practice it has four parts: the event you optimise toward (usually a trial or a payment rather than an install), the measurement setup that lets you see it, the channels you use and what each is for, and the creative process that feeds them.
What is a user acquisition campaign?
A paid campaign built to bring new users into an app, usually on Apple Ads, Meta, Google App Campaigns or TikTok. For subscription apps, a well-built UA campaign optimises toward trials or purchases rather than installs, so the platform finds users who pay rather than users who only download.
How much should an app spend on user acquisition?
There is no universal figure, but the discipline is: spend what your payback period can fund. If a cohort returns its cost in three months, you can sustain roughly three months of spend before it pays for itself, and that is what sets the ceiling. Starting budgets matter less than having one channel measured properly before you add a second.
Which user acquisition channel is best for a subscription app?
Apple Ads usually delivers the best cost per paying user because it captures existing intent, but it is capped by search volume. Meta provides the scale once the paywall and creative hold up. The practical answer for most subscription apps is to prove the funnel on Apple Ads, then scale on Meta, rather than choosing between them.
How do you measure user acquisition after ATT?
With a properly configured mobile measurement partner, a deliberate SKAdNetwork conversion schema rather than the default one, and where it suits the product, a web-to-app funnel that gives you first-party data before the store. Expect modelled rather than exact numbers on iOS, and build your decision process around confidence intervals instead of pretending the old precision still exists.
How can you reduce user acquisition costs?
Mostly by raising conversion rather than lowering media prices. Improving the store listing, onboarding and paywall lowers effective cost per acquisition across every channel at once. After that, move the optimisation event down the funnel, concentrate spend on the cohorts that already pay, and retarget installs that never converted.





