Why Your CAC Is Not Your CPI: How Game Studios Should Calculate Customer Acquisition Cost and Set an LTV:CAC Target
The first cost number on a game UA dashboard is CPI. But the number finance teams and investors ask about is CAC, customer acquisition cost, and the two are not the same. CPI is a media price: ad spend divided by installs. CAC is everything the company actually spent to bring one player in. Mistake a CPI that leaves out creative production, measurement tools, and UA team cost for CAC, and campaigns you believed were paying back can turn out to be losing money.
At current spend levels the gap is not small. AppsFlyer's State of Gaming for Marketers 2026, published in January 2026, puts global gaming UA spend at $25 billion in 2025, up 3.8% year over year, with 14.1 billion of the 24.8 billion game installs it analyzed coming from paid media. The same report found that top advertisers now run 2,400 to 2,600 creative variations per quarter, up 25% to 30% from a year earlier. The money spent outside media, on creative, people, and tools, is growing along with it.
We covered CPI benchmarks by genre and region in Mobile Game CPI Benchmarks 2026: Why One Average Is Useless. This post starts from that CPI and, as of September 2026, works through one hypothetical game end to end: how a game studio should define and calculate CAC, and how to set LTV:CAC and payback targets.
How is CAC different from CPI for a game studio?
CPI divides only media spend by installs; CAC divides every acquisition cost by the number of new users acquired. On top of that, a game studio needs to separate at least four versions of CAC depending on the denominator: paid users only or including organic, and installers or payers.
The general CAC formula is simple. HubSpot defines CAC as the combined cost of sales and marketing divided by the number of new customers, and the cost side includes not only ad spend but salaries, creative costs, software tools, and production. Translated to games:
CAC = (media spend + creative production + measurement and analytics tools + UA staff cost) ÷ new users
The denominator is where it gets interesting. The same costs answer very different questions depending on what you divide by.
Metric | Numerator | Denominator | Question it answers |
|---|---|---|---|
CPI | Media spend | Paid installs | Is the media price reasonable? |
Paid CAC (fully loaded) | Media + creative, tools, staff | Paid installs | What does one paid user really cost? |
Blended CAC | All acquisition costs | Paid + organic installs | What does the company spend per new user overall? |
CAC per payer | Media only or fully loaded | New payers | What does one revenue-generating user cost? |
Paid CAC and blended CAC serve different purposes. Because blended CAC puts organic installs in the denominator, it always comes out lower. By AppsFlyer's numbers, roughly 43% of game installs in 2025 (about 10.7 billion of 24.8 billion) were not paid, so the higher a game's organic share, the further blended CAC falls below paid CAC. That number is useful for company-level efficiency and for explaining unit economics to investors, but it is dangerous when deciding whether to raise ad spend, because incremental ad dollars only bring in paid users. The rule: use paid CAC for channel and campaign decisions and blended CAC for company-level reporting.
How do you actually calculate CAC? A worked example: a casual puzzle game spending $100,000 a month
The following is a hypothetical example for illustration. Assume a casual puzzle game spending $100,000 a month on US Android users; only the unit costs come from published sources.
Start with costs. If $100,000 in media buys 40,000 paid installs, CPI is $2.50, which sits inside Game Growth Advisor's 2026 range for US casual and puzzle Android CPI ($1.50 to $3.50). For creative, assume 20 video variations a month at $250 each, the midpoint of Game Growth Advisor's $150 to $350 per-variation range, for $5,000. Measurement is AppsFlyer's Growth plan rate of $0.07 per conversion applied to 40,000 paid installs, or $2,800 (AppsFlyer counts campaign-driven installs, re-engagements, and re-attributions as conversions and does not charge for unattributed organic installs). Analytics and creative tooling is assumed at $1,200 a month.
For staff, assume one UA manager dedicated to this game. According to the US Bureau of Labor Statistics, the median annual wage for advertising and promotions managers was $133,660 in May 2025, and as of June 2026 wages made up 70.0% of private-industry employer compensation costs, with benefits at 30.0%. Dividing the salary by 0.7 gives a fully loaded annual cost of about $191,000, or roughly $15,900 a month.
Cost item | Monthly amount | Basis |
|---|---|---|
Media | $100,000 | Assumption (40,000 paid installs at $2.50 CPI) |
Creative production | $5,000 | 20 variations × $250 (midpoint of $150–$350 range) |
MMP measurement | $2,800 | 40,000 paid installs × $0.07 |
Analytics and creative tools | $1,200 | Assumption |
UA staff | $15,900 | $133,660 median salary ÷ 0.7 ÷ 12 |
Total | $124,900 |
Add 20,000 organic installs in the same month, and assume payer conversion of 2.5% for paid users and 3.0% for organic users (these are assumptions; published payer conversion data is covered in the next section). The same month now produces five different numbers.
Metric | Calculation | Result |
|---|---|---|
CPI | 100,000 ÷ 40,000 | $2.50 |
Paid CAC (fully loaded) | 124,900 ÷ 40,000 | $3.12 |
Blended CAC | 124,900 ÷ 60,000 | $2.08 |
Paid CAC per payer (fully loaded) | 124,900 ÷ 1,000 payers | $124.90 |
Blended CAC per payer | 124,900 ÷ 1,600 payers | $78.06 |
Same game, same month, and the lowest number (blended CAC, $2.08) and the highest (paid CAC per payer, $124.90) are 60 times apart. Fully loaded paid CAC is 25% above CPI. The non-media share grows as ad spend shrinks, so for a small studio spending tens of thousands of dollars a month, the gap widens considerably. If an agency replaces in-house staff, its fees simply take the place of the salary line.
CAC per install or CAC per payer: which denominator should you use?
Games that earn mainly from ads should judge CAC per install; games that earn mainly from in-app purchases should judge CAC per payer. When payers are a few percent of the user base, a small change in payer conversion moves the price of a revenue-generating user far more than the install price does.
Published payer conversion data is scarcer than you might expect. AppsFlyer's The State of App Monetization 2026 (data from January 2025 to March 2026) reports that casino games convert 4.95% of new installs into paying users, with 3.01% making repeat purchases; non-gaming apps convert 9.84% of installs into first-time purchasers within 30 days, well above games. SolarEngine, a game analytics platform, describes typical first-purchase conversion in mobile games as 2% to 5%, with top titles at 5% to 8%. An older reference point: a 2019 GameAnalytics analysis of 100,000 games found most genres converting 1% to 2% of players into buyers, with arcade and word games below 1%.
CAC per payer is CAC divided by payer conversion, so it is highly sensitive to that rate. Holding the example's fully loaded paid CAC of $3.12 constant and changing only conversion:
Payer conversion | CAC per payer (at $3.12 fully loaded CAC) |
|---|---|
1% | $312 |
2% | $156 |
2.5% | $125 |
5% | $62 |
A drop of just half a percentage point, from 2.5% to 2%, raises CAC per payer by 25%. Raising payer conversion by half a point does more for CAC per payer than negotiating CPI down by 10%. That is why IAP-driven games should compare networks on CAC per payer, not CPI. A network with a low CPI that brings in users who rarely pay can end up with a higher CAC per payer.
For games like hyper-casual, where most revenue comes from ads, CAC per payer is meaningless. Compare CAC per install with ad revenue per install (ARPDAU times retained days). For hybrid monetization, track both denominators and make the one tied to the larger revenue share your primary metric.
What LTV:CAC ratio should a game target?
In games the right LTV:CAC ratio is not a single number. It depends on whether LTV is gross or net revenue and whether CAC means CPI or fully loaded cost, and each game should back-calculate its own threshold. SaaS's 3:1 rule comes from numbers defined differently and should not be copied directly.
The 3:1 rule comes from SaaS. In SaaS Metrics 2.0, David Skok writes that the best SaaS businesses have an LTV more than three times CAC, recover CAC in five to seven months, and that profitability becomes anemic when recovery stretches past twelve months. HubSpot likewise recommends a 3:1 CLV:CAC target. There are three reasons the rule does not transfer to games as is.
First, game LTV is usually quoted as gross revenue, before store fees, and store fees take a meaningful share of in-app purchase revenue. Apple's Small Business Program charges 15% to developers with up to $1 million in prior-year proceeds, reverting to the standard commission above that threshold. Google Play charges 15% on the first $1 million of annual earnings and 30% above it in most markets. From June 30, 2026, Google Play moved the EEA, UK, and US to a structure that separates new and existing installs and adds a separate 5% billing fee, so check which terms apply to your game. A 3:1 ratio on gross LTV is lower on net revenue.
Second, when games talk about LTV:CAC, the denominator is usually CPI. In SaaS, CAC is already the full sales and marketing cost; a game's LTV/CPI ratio leaves out creative, tools, and staff. The same 3:1 points at different things.
Third, game revenue is not contracted recurring revenue; it decays quickly after install. Because long-tail LTV is highly uncertain, studios often cap LTV at 180 or 365 days and manage against a payback rule such as "100% ROAS by day N" rather than a lifetime ratio.
Apply a 365-day LTV to the example game and four ratios emerge. Assume gross revenue per paid user over 365 days of $3.90 ($3.00 from IAP plus $0.90 from ads). Applying a 30% fee to IAP (an assumption) leaves net revenue of $3.00 ($2.10 IAP plus $0.90 ads). Ad revenue arrives after the ad network's share, so no store fee is applied to it.
Ratio definition | Calculation | Result |
|---|---|---|
Gross LTV ÷ CPI | 3.90 ÷ 2.50 | 1.56x |
Net LTV ÷ CPI | 3.00 ÷ 2.50 | 1.20x |
Net LTV ÷ fully loaded paid CAC | 3.00 ÷ 3.12 | 0.96x |
Net LTV ÷ blended CAC | 3.00 ÷ 2.08 | 1.44x |
On the ratio a dashboard typically shows, the campaign earns 1.56x. On a fully loaded basis, it has not recovered its cost after a full year (0.96x). Using blended CAC, it looks healthy at 1.44x, but justifying a budget increase with that number mistakes the cushion organic users provide for paid-user performance.
So how do you set your own threshold? If you want to keep using the dashboard's gross LTV/CPI, you can back-calculate the line it needs to clear:
Required gross LTV/CPI multiple = (gross revenue ÷ net revenue) × (fully loaded CAC ÷ CPI) × (1 + target margin)
For the example game, gross/net is 3.90 ÷ 3.00 = 1.30 and the fully loaded multiple is 3.12 ÷ 2.50 = 1.25. With a 20% target margin (an assumption), 1.30 × 1.25 × 1.20 ≈ 1.95x. In other words, this game needs a dashboard 365-day LTV/CPI above about 1.95x to recover all acquisition cost with a 20% cushion. At 1.56x today, it needs a lower CPI, a higher LTV, or a smaller non-media cost share. The point is to replace an external number like 3:1 with one that comes from your own cost structure. We covered the LTV side in How to Calculate LTV in Mobile Games for Smarter Growth, and whether LTV or CPI should anchor your UA plan in LTV vs CPI Strategy: What Should Drive Your Mobile Game UA Plan?.
How do you calculate CAC payback, and how many days should you target?
Payback is the day a user's cumulative net revenue first exceeds CAC. Because game revenue is not steady month to month, read it from the cumulative revenue curve rather than dividing. Set the target window from your cash position, not an industry average.
Subscription apps commonly use "payback months = CAC ÷ (monthly ARPU × gross margin)." Admiral Media's example is a paying user who costs $60 to acquire and pays $10 a month at 100% gross margin, for a six-month payback, and it describes six to twelve months as the healthy range for subscription apps. But game revenue is concentrated in the first week and month and falls off quickly afterward, so a formula that assumes the same amount every month will misread payback.
Assume the example game's cumulative net revenue per paid user follows this curve (illustrative):
Day | Cumulative net revenue | vs CPI ($2.50) | vs fully loaded CAC ($3.12) |
|---|---|---|---|
D7 | $0.55 | 22% | 18% |
D30 | $1.25 | 50% | 40% |
D90 | $2.00 | 80% | 64% |
D180 | $2.60 | 104% | 83% |
D365 | $3.00 | 120% | 96% |
Against CPI, payback lands between D90 and D180. Against fully loaded CAC, it does not arrive within 365 days. The same campaign gets opposite verdicts, "pays back within six months" versus "not paid back after a year," depending on which CAC you use.
Payback within 30 days is rare in games. Liftoff's 2025 Casual Gaming Apps Report found average D30 ROAS for casual games in 2024 of 47% on iOS and 15% on Android, and 60% and 27% for strategy games. When the average campaign recovers less than half its spend in the first month, payback targets end up measured in months.
The constraint that sets the window is cash. Writing in Gamigion, Sam Nam uses a business constraint as the starting point for a ROAS target: given the bank balance and the app's payment cycle, all UA campaigns need to reach 100% ROAS within 120 days. A studio with little cash has to shorten the window and give up scale; one with more room can lengthen it and buy more expensive users. Once the window is set, you can back-calculate interim D7 and D30 targets, as described in D7 ROAS Benchmarks: The Targets That Tell You a Campaign Will Pay Back. Set those interim targets against fully loaded CAC rather than CPI to avoid the trap in the table above.
How do CAC baselines differ by genre?
Changing genre changes CPI levels, monetization model, and payback speed together, so the CAC denominator and targets change too. Ad-driven genres default to CAC per install and short payback; payer-driven genres default to CAC per payer and longer payback.
The table below uses Admiral Media's 2026 tier-1 iOS CPI ranges and D90 ROAS target ranges by genre. Admiral Media notes that Android CPIs typically run 40% to 60% below iOS for the same genre.
Genre | Tier-1 iOS CPI | D90 ROAS target | Primary CAC denominator |
|---|---|---|---|
Hyper-casual (ad-driven) | $0.80–$2.50 | 110–150% | Installs |
Casual (IAP + ads) | $2.50–$6.00 (casual puzzle) | 90–140% | Installs + payers |
Idle RPG | $4.00–$9.00 | Not available | Payers |
Mid-core / strategy | $8.00–$18.00 | 70–120% | Payers |
Slots / social casino | $20.00–$60.00 | Not available | Payers |
The D90 target is highest for hyper-casual and lowest for mid-core because of payback speed. Ad-driven games lose users quickly, so they must be paid back inside 90 days; in mid-core, payers stay longer and revenue keeps accruing after day 90, so a campaign can still be short of payback at that point.
Converting to CAC per payer widens the gap. For casino, applying AppsFlyer's 4.95% payer conversion gives an iOS media-only CAC per payer of about $404 to $1,212. For other genres there is no published genre-level payer conversion data, so using assumptions: mid-core and strategy at 3% gives about $267 to $600, and casual puzzle at 2% gives about $125 to $300. Add creative, tools, and staff and fully loaded CAC per payer rises further. This is why a cross-genre average CAC is useless. The starting point for any comparison is your own genre's CPI range combined with your own game's payer conversion.
If your CAC comes in above target, we have outlined the levers in How to Reduce User Acquisition Cost with Proven Marketing Strategies.
What changes when engaged players, not installs, are the denominator: the Playio perspective
Move the CAC denominator from installs to users who actually entered and played the game, and the cost hidden between CAC per install and CAC per payer, users who installed but never played, becomes visible. Put the pricing itself on post-install behavior, and that cost can be cut at the point of choosing a network.
As the sensitivity table showed, the biggest driver of CAC per payer is not CPI but the share of installers who go on to pay, and that share depends heavily on whether users get into the game's core loop in the first few days. A network with cheap installs but heavy early churn produces expensive payers.
Playio runs an Android-based community of five million gamers, matching games to players using genre preference, play history, and in-game behavior data, and tying rewards to in-game actions such as reaching a playtime threshold or a specific level. Pricing can run on CPI or CPE. On CPE, spend goes only to users who completed the target action rather than to installs, so "CAC per engaged user" becomes the contracted price. Because the cost basis moves to the step just before payer conversion, the math gets simpler for payer-driven genres that have to manage CAC per payer.
You can find more details here. (https://playioadsen.oopy.io/bizdeck)
Key Takeaways
As of September 2026, the first step in managing CAC is separating it from CPI. CPI is a media price; CAC is the full cost of acquisition, including creative production, measurement tools, and staff. In the hypothetical example, fully loaded paid CAC was 25% above CPI, and the same month's CAC ranged 60-fold, from $2.08 to $124.90, depending on the denominator. Use paid CAC for channel decisions and blended CAC for company-level reporting, and compare networks on CAC per payer for IAP-driven games. Rather than borrowing SaaS's 3:1, back-calculate your own threshold multiple from store fees, non-media costs, and target margin. The example game showed 1.56x on the dashboard but 0.96x on a fully loaded basis, failing to pay back within a year, against a threshold of about 1.95x. Read payback from the cumulative revenue curve, and set the target window from your cash position.
For inquiries about Playio's advertising solutions, reach out at: [email protected]
Sources
AppsFlyer, State of Gaming for Marketers 2026, January 14, 2026 ($25B gaming UA spend in 2025, +3.8%; 14.1B paid of 24.8B installs; top advertisers at 2,400–2,600 creatives per quarter, +25–30%; 9,600 gaming apps analyzed): https://www.appsflyer.com/company/newsroom/pr/gaming-marketing/
HubSpot, What Does CAC Stand For?, updated October 2, 2025 (CAC = (cost of sales + cost of marketing) ÷ new customers; cost categories; 3:1 CLV:CAC recommended): https://blog.hubspot.com/service/what-does-cac-stand-for
Game Growth Advisor, Mobile Game KPIs 2026 (US casual/puzzle Android CPI $1.50–$3.50; notes that no reliable published payer conversion benchmark was found): https://gamegrowthadvisor.com/blog/2026-03-17-mobile-game-kpis-benchmarks-2026/
Game Growth Advisor, Mobile Game Ad Creative Strategy 2026 (video variations on an established concept at roughly $150–$350 each): https://gamegrowthadvisor.com/blog/2026-05-12-mobile-game-ad-creative-strategy-2026/
AppsFlyer Pricing (Growth plan $0.07 per conversion after the welcome package; unattributed organic installs free): https://www.appsflyer.com/pricing/
US Bureau of Labor Statistics, Occupational Outlook Handbook (median annual wage for advertising and promotions managers $133,660, May 2025): https://www.bls.gov/ooh/management/advertising-promotions-and-marketing-managers.htm
US Bureau of Labor Statistics, Employer Costs for Employee Compensation, June 2026 (private industry wages 70.0%, benefits 30.0%): https://www.bls.gov/news.release/ecec.nr0.htm
AppsFlyer, The State of App Monetization 2026 (January 2025–March 2026; casino games convert 4.95% of new installs to payers, 3.01% repeat; non-gaming apps 9.84% first-time purchasers within 30 days): https://www.appsflyer.com/resources/reports/app-marketing-monetization-report/ (summary: https://gamedevreports.substack.com/p/appsflyer-app-monetization-in-2026)
SolarEngine, From Player to Payer (typical first-purchase conversion 2–5%, top titles 5–8%): https://blog.solar-engine.com/en-blog/docs/From-Player-to-Payer-The-Guide-to-Cracking-FirstPurchase-Conversion-in-Mobile-Games
AppFollow, Mobile game analytics with Ioana Hreninciuc (2019 GameAnalytics analysis: most genres 1–2% conversion to buyers, arcade and word below 1%): https://appfollow.io/blog/mobile-game-analytics-with-ioana-hreninciuc
David Skok, SaaS Metrics 2.0 (LTV above 3x CAC; CAC recovered in 5–7 months; anemic profitability beyond 12 months): https://www.forentrepreneurs.com/saas-metrics-2/
Apple Developer, App Store Small Business Program (15% for developers with up to $1M in prior-year proceeds; standard rate above): https://developer.apple.com/app-store/small-business-program/
Google Play Console Help, Service fees (most markets 15% on first $1M annual earnings, 30% above, 15% on auto-renewing subscriptions; from June 30, 2026, new vs existing install tiers plus a 5% billing fee in the EEA, UK, and US): https://support.google.com/googleplay/android-developer/answer/112622?hl=en
Admiral Media, CAC Payback Period for Mobile Apps (payback months formula; $60 / $10-per-month example; 6–12 months for subscription apps): https://admiral.media/cac-payback-period-mobile-apps/
Liftoff, Must-Know Highlights From the 2025 Casual Gaming Apps Report (2024 casual D30 ROAS 47% iOS, 15% Android; strategy 60% iOS, 27% Android): https://liftoff.ai/blog/highlights-2025-casual-gaming-apps-report/
Sam Nam, What Should Your ROAS Target Actually Be?, Gamigion (120-day 100% ROAS target derived from cash balance and payment cycle): https://www.gamigion.com/what-should-your-roas-target-actually-be/
Admiral Media, Mobile App Marketing Benchmarks 2026 (tier-1 iOS CPI ranges by genre; Android 40–60% lower; D90 ROAS targets by genre): https://admiral.media/mobile-app-marketing-benchmarks-2026/
The example game's install counts, payer conversion (2.5% paid, 3.0% organic), tool cost, 365-day LTV and cumulative revenue curve, 30% IAP fee, 20% target margin, and the 3% mid-core and 2% casual payer conversion rates are illustrative assumptions; CAC per payer and ratios are our own calculations from the figures above