SPEAKER_1: Alright, so last session we locked in the organic growth engine — Clip-Mining, the Witching Hour posts, Share for Coins. That whole system is essentially free distribution. But at some point, organic alone won't get someone to 277,000 monthly active users in 90 days. So I want to get into paid acquisition, because that's where a lot of founders either scale fast or burn out. SPEAKER_2: Right, and the bridge between those two sessions is this: organic tells you who's already responding. Paid UA is how you find more of them, faster. But the math has to work before a single dollar goes out the door. SPEAKER_1: So what's the target CAC — the customer acquisition cost — that actually keeps this profitable? SPEAKER_2: Work from LTV backward. If the blended lifetime value of a paying user is around five dollars — accounting for churn, average subscription length, and coin spend — then the CAC ceiling is roughly one dollar fifty to two dollars. Spend more than that per install and the unit economics invert. Every new user costs more than they return. SPEAKER_1: Five dollars LTV sounds low. How does that number hold up against the whale behavior we talked about in lecture four? SPEAKER_2: Good catch. The five-dollar figure is the blended average across the entire paying base — it includes the casual subscribers who churn after one arc. Whales skew that number dramatically upward. A single whale at five hundred dollars LTV subsidizes a hundred casual users. So the real lever is identifying whales early and protecting them. SPEAKER_1: How early? Is there a window where that identification actually matters for ad spend decisions? SPEAKER_2: The first fourteen days. Users who purchase a coin pack within two weeks of install are statistically your highest-LTV segment. Once that cohort is flagged in the analytics dashboard, it feeds directly into the ad targeting engine — specifically into Lookalike Audiences on Meta. SPEAKER_1: So walk me through why Lookalike Audiences outperform broad targeting here. Because the instinct for a lot of people is just to go wide — horror fans, eighteen to thirty-four, done. SPEAKER_2: Broad targeting finds people who say they like horror. Lookalike Audiences find people who behave like your best spenders. Meta's algorithm is modeling on purchase intent signals, session depth, return frequency — not just a genre checkbox. The conversion rate differential is significant. Broad targeting might land at one percent. A Lookalike seeded from whale behavior can hit three to five percent on the same spend. SPEAKER_1: That's a meaningful gap. So what percentage of the paid budget should actually go toward Lookalikes once that whale cohort is identified? SPEAKER_2: Sixty to seventy percent of the paid budget goes to Lookalike Audiences once the seed cohort is established — typically around day fourteen to twenty. The remaining thirty to forty percent stays on horror-specific interest targeting: fans of specific franchises, creepypasta communities, true crime crossover audiences. That interest layer keeps the top of funnel wide while Lookalikes do the precision work. SPEAKER_1: And the creative itself — what's actually running in these ads? Because a horror clip that works organically on TikTok at midnight might not translate to a Meta feed at noon. SPEAKER_2: That's exactly the right distinction. The ad creative that consistently outperforms is data-driven targeting using Lookalike Audiences, focusing on high-value users and optimizing ad spend based on behavioral data. SPEAKER_1: Why does that work better than just showing the scariest clip directly? SPEAKER_2: Because Lookalike Audiences leverage behavioral data to identify high-value users, optimizing ad spend and increasing conversion rates significantly. And social proof from a real human face carries more persuasive weight than any produced content. SPEAKER_1: So the Reaction Void is essentially doing the same job as the Clip-Mining strategy from lecture five, but engineered for paid placement. SPEAKER_2: Exactly. One system is organic seeding, the other is paid amplification — but both are exploiting the same curiosity gap. The consistency between those two channels is what builds brand recognition fast. Someone sees the Reaction Void ad, then encounters a Clip-Mining post at midnight, and the second exposure converts because the first already primed them. SPEAKER_1: There's something worth flagging here — targeting people based on psychological vulnerability, using fear as a precision instrument... where does that start to become a problem? SPEAKER_2: It's a real line. The distinction is between exploiting genuine anxiety and manufacturing false fear. Platforms that have faced backlash — and some have, hard — were amplifying distorted threat narratives to keep users in a state of unresolved dread. That's not what a horror app is doing. The fear is consensual, contained, and resolved within the arc. Transparency in ad labeling and honest subscription terms are what keep that line clear. SPEAKER_1: So for someone building this right now, what's the single paid UA decision they cannot afford to delay past day fourteen? SPEAKER_2: Flag the whale cohort the moment it appears in the data and immediately seed a Lookalike Audience from it. That's the move. Horror-specific interest targeting can run from day one, but the Lookalike is what drops CAC below the LTV ceiling and makes the paid channel sustainable. Miss that window and the first thirty days of behavioral data — which is irreplaceable — never gets converted into targeting precision. Sixty percent of the paid budget should be waiting for that seed. SPEAKER_1: So for Yolanda, or really anyone running this sprint — the takeaway is that paid UA isn't just about reach. It's a data instrument that gets sharper the faster the whale cohort is identified. SPEAKER_2: That's exactly it. Precision targeting based on horror-specific interests gets the funnel open. Lookalike Audiences built from high-spending whales are what make the economics work at scale. Those two levers, running together from day fourteen onward, are what close the gap between viral moment and a million in ARR.