PILLAR ESSAY · B2C SAAS
Your dashboard is green. Installs are up. Cost-per-install is down. Paying subscribers are flat.
The algorithm is optimizing for the event you told it to find. The event you told it to find is install. Installers and paying subscribers are different audiences.
Here is the optimization-event diagnostic that fixes the gap, plus the 5-check framework I run on every B2C SaaS audit.
B2C SaaS Paid Strategy: Why Cost-Per-Install Lies
By Eleni Buras | Published 2026-05-15 | Updated 2026-05-16
TL;DR
- Cost-per-install is the wrong number. The algorithm optimizes for what you ask it to find, and install is the easiest event to find.
- The optimization-event mismatch is the single most common B2C SaaS paid failure mode I see in audits.
- The fix is to optimize against a paid-subscription event: trial start, first paid month, or a server-side revenue signal.
- Target paying-subscriber CAC payback under 9 months. Cohort retention by channel from week one.
- Same vocabulary as DTC. Different math. Applying DTC strategy by default is why most B2C SaaS founders underperform.
Your dashboard looks healthy.
Installs are growing month over month. Cost-per-install is dropping. The creative team is shipping new angles. The agency report shows green across the board.
Your paying subscribers are not growing.
This is the most common failure mode I see in B2C SaaS paid programs at seed and Series A. The platform algorithm is optimizing for the event you told it to find. The event you told it to find is install. Installs are getting cheaper because the algorithm is getting better at finding installers. Installers are not paying subscribers.
I have audited dozens of B2C SaaS subscription apps across health, wellness, productivity, finance, dating, language learning, and parenting. The pattern repeats. The fix is technical, cheap, and almost always overlooked by the founder and the agency until someone outside the program names it.
This is the long-form answer to the question I get on every B2C SaaS discovery call: why is the cost-per-install great and the paid subscriber growth flat.
The optimization-event mismatch
The mismatch is the source of the problem. Understanding it is the entire game.
When you run paid acquisition on Meta, TikTok, Apple Search Ads, or Google App Campaigns, you are telling the platform what event you want it to optimize toward. The platform’s algorithm then searches for users statistically similar to people who took that event.
If you tell the platform to optimize for install, the algorithm finds users statistically similar to people who install apps. App-installers as an audience are abundant, casual, and frequently uninterested in paying. The pool is enormous and the cost-per-install is low.
If you tell the platform to optimize for paid trial start, the algorithm finds users statistically similar to people who start paid trials. The pool is smaller. The cost-per-event is higher. But the audience converts to paid subscription at a fundamentally different rate.
The numerical impact is dramatic. In audits I have run, switching the optimization event from install to paid trial start typically moves free-to-paid conversion from 3 to 6 percent up to 8 to 14 percent within 60 to 90 days. The install cost goes up. The cost-per-paying-subscriber goes down. The dashboard looks worse in week one and dramatically better in week eight.
Most founders and most agencies optimize for install because:
- It is the default setup the platform suggests
- It produces the highest event volume, which the algorithm needs to learn against
- It produces the cheapest cost-per-event, which looks good in reports
- It is the easiest event to track without server-side infrastructure
None of those reasons are wrong on their own. They are wrong in combination. The first three create the appearance of a healthy program. The fourth is why the technical fix is overlooked.
Why the platform’s default is wrong for subscription
The platforms (Meta, TikTok, Apple Search Ads) are not neutral. Their defaults reflect what their algorithm performs best on, which is high-volume top-of-funnel events. For app marketing, that is install.
The default makes sense for advertisers who actually want installs. Games, free utility apps, content apps with ad-based monetization, lead-gen apps. For subscription apps, the default is misaligned with the business model.
Subscription apps make money on retained paying users. The economics are determined by what happens 14, 30, 60, and 90 days after install. The algorithm optimizing for install has no information about that.
Telling the platform that paid trial start is the goal pulls the algorithm into a different optimization space. It learns from a different conversion signal. It finds users with statistically different intent. The audience composition shifts.
This is not a small adjustment. It is the difference between buying installs (which you do not need) and buying paying subscribers (which you do).
Defined terms (subscription-app specific)
Optimization event. The conversion the platform algorithm is told to find. The event you optimize for shapes the audience the platform brings you. Wrong event, wrong audience, wrong dashboard.
Free-to-paid conversion (FTP). The percentage of free signups who become paying subscribers within a cohort window, typically 30 or 60 days. Healthy subscription app FTP at scale is 8 to 15 percent. Below 5 percent usually signals an optimization-event mismatch or a product onboarding problem.
Paying-subscriber CAC. Total paid spend divided by new paying subscribers in the same period. This is the only CAC number that matters for subscription apps. Install-cost CAC is not a real metric.
Trial start. A free or paid trial activation event. Far enough down-funnel to filter casual installers. Close enough to install to provide volume the algorithm can learn against.
Server-side event tracking. Sending conversion events to the platform from your own server (not from the device), so that platform attribution survives iOS privacy changes, SKAdNetwork limitations, and cross-device user journeys. Conversion API for Meta, Events API for TikTok, MMP integration for Apple Search Ads.
60-day churn by channel. Percentage of paying subscribers acquired in a given month who churn within 60 days, segmented by acquisition channel and creative angle. The single most predictive number for long-term subscription economics.
ARPU. Average revenue per user. For subscription apps, usually monthly ARPU at scale. Used in CAC payback calculations.
The 5-step B2C SaaS paid diagnostic
This is the framework I use in B2C SaaS audits. Five checks, in order. Skipping any of them is what produces the “installs are great, subs are flat” dashboard.
Check 1: What event is the platform optimizing for
Open the platform manager. Find the active campaigns. Look at the conversion event each ad set is optimizing toward.
If the event is “install,” “app install,” or “mobile app install,” you have the mismatch. Move to Check 2 to plan the fix.
If the event is “trial start,” “first paid month,” or a custom revenue event, you are optimizing correctly. Move to Check 3.
If the event is “registration” or “free signup,” you are in a middle zone. Better than install for most subscription apps, worse than trial start. Decide based on FTP volume and how mature your activation funnel is.
Check 2: Is the server-side event infrastructure in place
The optimization-event fix only works if the platform actually receives the paid-subscription signal. Most apps rely on the platform’s SDK to capture conversion events, which works inconsistently because of iOS privacy changes and SKAdNetwork limitations.
Server-side event tracking sends the events directly from your server to the platform’s API. The setup is technical but cheap. Most subscription infrastructure tools (RevenueCat, Adapty, Glassfy) have one-click integrations for the major platforms.
If the server-side setup is missing, the fix is to add it. Two to four weeks of engineering time for the integration and testing.
Check 3: What is the 60-day churn by acquisition channel
Pull the cohort table. Acquisition month on one axis, retention curve on the other, broken down by channel.
The pattern you are looking for: which channels produce subscribers who stay versus which produce subscribers who churn at 60 days. Most subscription apps find that one or two channels are doing the work and the rest are paying for installs that churn.
The fix is to reallocate spend from the high-churn channels to the high-retention channels, then refresh creative on the high-retention channels to maintain volume.
Check 4: What is the creative angle library doing
Subscription apps usually have one or two creative angles in heavy rotation, plus a tail of weaker variations. The heavy-rotation angles drive most of the volume but also saturate fastest.
The fix is a buyer-state-matrix approach. Map angles against the user’s awareness stage (problem-aware, solution-aware, brand-aware) and against the emotional driver (relief, identity, status, ritual). For a wellness app: relief angles for problem-aware users with sleep issues, identity angles for solution-aware users comparing apps, ritual angles for brand-aware users considering renewal.
Three to five new angles per quarter, rotated against the existing library, keeps the audience fresh and the cost-per-paying-subscriber stable.
Check 5: What does the LTV curve actually look like
Pull the LTV-by-cohort table for the last 6 to 12 months. Look at the actual LTV at month 6, month 9, month 12. Compare against your CAC.
If you do not have an LTV table by cohort, you are guessing at unit economics. Most subscription apps overestimate LTV because they use the optimistic curve (paying subscribers who stay) instead of the blended curve (all paid subscribers, including those who churn).
The fix is to rebuild the LTV calculation using actual cohort retention, then recalculate CAC payback honestly. This often reveals that the program was deploying paid against an LTV number that does not exist.
Five checks. The hardest two are server-side and cohort.
The 5-check diagnostic above is something you can run yourself. The hardest pieces are the server-side event setup (2 to 4 weeks of engineering) and the cohort retention analysis by channel (data infrastructure most subscription apps have not built yet).
The 10-Day Allocation Audit covers the diagnosis, the channel reallocation, the creative-angle matrix, and the 90-day plan. $4K fixed fee. Strategy only. We do not run your ads. One in five audits ends with a wait recommendation.
The 5-step diagnostic in practice
A walk-through from a recent engagement.
The company. $2M ARR seed-stage consumer subscription, mental wellness category. Monthly subscription $9.99 or annual $59.99.
Initial dashboard. Cost-per-install $2.80 on Meta, $4.10 on TikTok. Install volume up 40 percent quarter over quarter. Free-to-paid conversion sitting at 4 percent. Paying subscribers up 6 percent quarter over quarter. Founder confused about the gap between install growth and subscriber growth.
Check 1. Both Meta and TikTok campaigns were optimizing for “Mobile App Install.” Mismatch confirmed.
Check 2. Server-side event infrastructure was partially in place via RevenueCat but not piping trial-start events to the platforms. Two-week engineering sprint to complete the setup.
Check 3. 60-day churn by channel was 78 percent on TikTok and 52 percent on Meta. TikTok was producing cheap installs that churned. Meta was producing more expensive installs that stayed.
Check 4. Three creative angles in rotation, all problem-aware (“can’t sleep” framing). No identity or ritual angles addressing solution-aware or brand-aware users.
Check 5. LTV calculation assumed 12-month retention. Actual blended cohort retention at 12 months was 38 percent. Real LTV roughly half of the planning number.
The fix. Switched optimization event to paid trial start on both Meta and TikTok. Completed server-side event setup. Reallocated 30 percent of TikTok budget to Meta where retention was twice as strong. Added four new creative angles addressing solution-aware and brand-aware user states. Rebuilt LTV planning model with actual cohort data.
Result at 90 days. Free-to-paid conversion moved from 4 percent to 11 percent. Paying-subscriber CAC dropped from roughly $90 to roughly $40 blended. 60-day churn dropped from 78 percent to 41 percent on TikTok at the new optimization event. Paying-subscriber growth tripled quarter over quarter at flat spend.
This is what the diagnostic produces. Not always with these specific magnitudes, but the pattern is the same.
Comparison: B2C SaaS paid vs DTC paid vs B2B SaaS paid
The vocabularies overlap. The math does not.
| B2C SaaS Subscription | DTC E-commerce | B2B SaaS | |
|---|---|---|---|
| Primary conversion event | Paid trial start or first paid month | Purchase | Demo booked or SQL |
| Right optimization event | Trial start (server-side) | Purchase (server-side, Conversion API) | Demo or qualified form fill |
| CAC measurement | Paying-subscriber CAC | Order-level CAC, then customer-level | Closed-won deal CAC |
| Payback target | Under 9 months | Contribution-margin positive on first order, repeat within 90 days | Under 14 months |
| LTV reality | Cohort retention curve unfolds over 6 to 24 months | Repeat purchase pattern unfolds over 30 to 180 days | Expansion revenue over 12 to 36 months |
| Channel mix | Meta primary, TikTok creative, Apple Search Ads intent | Meta primary, TikTok angle dev, Google Shopping | Intent search primary, LinkedIn ABM-light, retargeting |
| Creative cadence | 3 to 5 new angles per quarter against buyer-state matrix | 60 to 90 day creative refresh per channel | Slower, depth over breadth |
| Most common failure mode | Optimization-event mismatch (install vs paid sub) | Creative saturation in angle library | Equal-thirds channel split before motion is mapped |
| Attribution window | 7 to 30 days from install to first paid event, then ongoing | 1 to 14 days from click to purchase | 30 to 90 days from first touch to closed-won |
The differences are why applying DTC paid strategy to a B2C SaaS app by default is the most common over-confident mistake at seed and Series A. The shared vocabulary creates a false sense of transferability.
What B2C SaaS founders get wrong by default
A short list, in order of frequency.
Optimizing for install or registration. Covered above. The single biggest fix.
Comparing cost-per-install to industry benchmarks. The benchmark is meaningless if the optimization event is wrong. Cost-per-paying-subscriber is the only useful number.
Running the same creative angles for six months. The audience inside a single emotional driver saturates. Three to five new angles per quarter against the buyer-state matrix.
Treating Apple Search Ads as an afterthought. ASA captures the highest-intent traffic in the funnel. Allocating 10 to 15 percent of budget to ASA against branded and competitor terms is usually the highest-ROAS portion of the program at scale.
Underinvesting in server-side attribution. Two weeks of engineering work that pays back inside 60 days through better algorithm targeting.
Reporting install volume to the board. Investors increasingly know the optimization-event problem. Reporting paying-subscriber growth, free-to-paid conversion, and 60-day cohort retention is the credible board metric stack.
Hiring a generalist performance agency. B2C SaaS subscription paid is different enough from DTC and B2B that most generalist agencies underperform. Specialist subscription-focused execution is worth the premium.
When to bring in outside help
The 5-step diagnostic above is something you can run yourself. The hardest parts are the server-side event setup and the cohort analysis, both of which require some technical investment.
The reason B2C SaaS founders bring me in is usually one of three. Free-to-paid conversion is stuck below 6 percent and they suspect the optimization event but have not been able to prove it. They are about to raise a round and need a diligence-grade paid plan that addresses the install-vs-paying-sub question directly. Or 60-day churn by channel is bad and they need to know whether to reallocate or refresh creative.
The 10-Day Allocation Audit covers all three. $4K fixed fee. The deliverable includes the optimization-event diagnosis, the cohort analysis, the channel reallocation recommendation, the creative-angle matrix, and the 90-day spending plan. Strategy only. We do not run your ads.
Frequently Asked Questions
Why are my installs growing but paid subscribers are not?
Because the platform algorithm is optimizing for the event you told it to find, and the event you told it to find is install. The algorithm is excellent at finding people who install apps. It is not finding people who pay for subscriptions. The fix is to change the optimization event to a paid-subscription signal: trial start, first paid month, or a server-side revenue event. The algorithm will then find a different audience.
What is the right optimization event for a B2C SaaS subscription app?
Paid trial start is the cleanest event for most subscription apps. It is far enough down-funnel to filter casual installers, close enough to install to give the algorithm enough volume to optimize against. For mature apps with stable trial-to-paid conversion above 25 percent, optimizing for first paid month is even better. The wrong event is install, even though it is the easiest to set up.
What CAC payback should a B2C SaaS subscription target?
Paying-subscriber CAC payback under 9 months at scale, 6 months for best-in-class. Calculation: paying-subscriber CAC divided by (monthly ARPU times gross margin times expected retention curve). Subscription apps often look like they have 3-month CAC payback when measured against install cost. That number is meaningless. Measure against paying-subscriber CAC.
How do I fix attribution for a subscription app?
Server-side event tracking is the minimum. Pipe revenue events (trial start, first paid charge, second paid month) back to Meta, TikTok, and Apple Search Ads via their server-side APIs. Use RevenueCat or a similar subscription infrastructure layer to deduplicate. Do not rely on platform-attributed installs; the attribution windows are too short for the time-to-pay reality of most subscription apps.
What is the right channel mix for a B2C SaaS subscription at Series A?
Meta carries 50 to 60 percent because it can optimize for paid-subscription events when the server-side setup is correct. TikTok carries 15 to 25 percent for creative-angle development at the top of funnel. Apple Search Ads carries 10 to 15 percent for high-intent capture of branded and competitor terms. Retargeting carries the balance. Adjust based on cohort retention by channel, not by install volume.
When should I worry about 60-day churn versus first-day install?
Immediately. 60-day churn by acquisition channel is the single most predictive number for the long-term economics of a B2C SaaS subscription. A channel that produces installs at $4 with 80 percent churn at 60 days is worse than a channel that produces installs at $8 with 35 percent churn. The CAC math punishes the high-churn channel even though it looks cheaper. Cohort by channel and by creative angle from week one.
How is B2C SaaS paid strategy different from DTC e-commerce paid strategy?
The optimization event is different (paid subscription vs purchase), the attribution window is different (subscription LTV unfolds over months, DTC purchase is immediate), the creative strategy is different (subscription apps fight for ongoing usage, DTC fights for transaction). The vocabulary overlaps (CAC, ROAS, creative angles) but the math is different. Most B2C SaaS founders apply DTC paid strategy by default and underperform because the math does not transfer.
Install volume is not subscriber growth.
If the gap on your dashboard looks familiar, the audit is the document that closes it. Optimization-event diagnosis, server-side event spec, cohort analysis by channel, creative-angle matrix, 90-day spending plan. $4K fixed. 10 days. Strategy only.
One in five audits ends with a wait recommendation, which for a subscription app usually means "your activation funnel is the bottleneck, not paid." Worth knowing before another $50K of installs comes in.
About the author
Eleni Buras is the founder of EBP Digital, a paid media allocation consultancy for newly-funded B2B SaaS, B2C SaaS, and e-commerce/DTC founders. Ex-Marketing Director with 20 years deploying ad budgets at funded SaaS and e-commerce brands. $100M+ in paid programs audited. The 10-Day Allocation Audit covers the optimization-event diagnostic, the cohort analysis, the channel reallocation, and the 90-day plan. Strategy only. We do not run your ads.
Book the 10-day allocation audit.
Related pillars:
- Where to Spend the First $100K of Ad Budget After You Raise
- How to Allocate Paid Media Budget After Series A
- The Paid Media Plan That Holds Up in Diligence
- Fractional CMO vs Paid Media Strategist vs Agency
- ROAS Dropped From 3.5x to 1.8x: The DTC Diagnostic Flowchart (coming)