Attribution and signal tactics

Use only the Meta SDK at the start. Send install, paywall view (standard View Content event, fires when the paywall loads), initiate checkout, start trial and subscribe. In Meta Events Manager > Data sets > your app, pick the same date range as App Store Connect. Compare installs with App Store Connect Analytics > Total Downloads, and start trials with Analytics > Metrics > Subscriptions > Events > Offer Starts filtered to Free Trial Starts. Over Sept 14 - Oct 10: ~3,000 downloads in App Store Connect vs a little over 4,000 in Meta; 312 trial starts vs 276 in Meta.

Register a benefit-named domain used only in the agency's organic videos (flightcheatcode.com for Jetpack; buyabusiness.com / buythebusinessyouwant.com for SMB Market) that redirects straight to the app/App Store. Separates organic creator traffic from paid ads, other agencies and in-house creators running at the same time, and is easier to remember than an unknown brand name. Used only for brands without a well-known or viral name (not needed for e.g. Polymarket). Tested on a couple of apps; SMB Market's domain went live a few weeks before the episode.

Context: on Meta, credit/fintech ads can't use lookalikes, age, gender, ZIP or specific interests. Funnel was install -> sign-up -> KYC -> bank link -> first transfer (revenue event). Sign-up was too loose (people gaming cashback/referral rewards); first transfer was too deep (three high-friction steps). At sign-up, data was routed to the internal database and the first and last name run through a name-origin API; anyone with 70%+ probability of being from South Asia (the Indian expat target) triggered a synthetic 'HQ' event sent back to the ad network. No financial data or PII was sent. The event met the three criteria: enough volume (Facebook 50/7 days, Google 10/campaign/day), recency (within 24-48h) and a good proxy for first transfer. Guest notes such scoring could now be vibe-coded instead of using an API.

For a 7-day free trial that auto-renews into a subscription, analyzed how soon after starting the trial people cancel: 25-30% cancelled within the first 12 hours. Instead of firing trial-started immediately, sent an adjusted trial-started event to Meta after ~12 hours (host: '12 hours and 5 minutes') only for users who hadn't cancelled, so the algorithm finds users more likely to pay. Done around 2023 or earlier; guest says it was an industry first then and is now commonplace.

Once an organic video passes ~5% engagement rate, put spend behind it on Meta (SideShift runs Meta primarily, for its B2B product). Start around $100/day total across the winners (e.g. ~$33 each for 3 videos) for the first 14 days, adjusted for trial length. Track trial starts and conversions via UTM links. If ROAS is positive (e.g. $200 in, $400 out), step up to $300/day, then $1,000/day; if you get 75 cents per dollar, stop. Start paid only after organic content-market fit and once total marketing budget exceeds ~$10K/month.

Compute RPM (revenue per 1,000 views) from past videos: ~$3-4 for Locked (~10% download-to-sub, $30/yr or $7/wk). Offer CPM deals of $1-2 per 1,000 views, always with a payout cap (e.g. $500) set upfront, so paid CPM stays under RPM.

Campaigns still optimize for purchase, but the app also sends Meta's standard e-commerce events: fb_mobile_complete_registration on sign-up, fb_mobile_content_view once when the paywall is shown, and fb_mobile_initiated_checkout once when the paywall buy button is tapped. Mapped via Adjust (works with the plain Facebook SDK too). Meta then learns from registration, paywall view and checkout intent before the first purchase arrives; cost per checkout also shows which ads bring quality traffic.

In Adjust: fine conversion values ordered install -> registration -> paywall view -> initiated checkout -> subscription revenue tiers by monthly/quarterly/annual (annual highest; threshold set at $43 USD so $59 CAD annual counts). Coarse values for the first postback (0-2 days): started with paywall view as 'medium' while purchases were scarce; at 100+ installs/day switched medium to any purchase and high to annual. Second (3-7 days) and third (8-35 days) postbacks: medium = monthly/quarterly, high = annual.

Turned off all default Meta app events and send a single purchase event worth 25 EUR whenever a user starts a free trial, buys monthly, or takes the 50% offer, with no 7-day wait. Value set below the ~40 EUR blended LTV per converting user. Each week, check that actual trial -> paid matches the projection in the P&L sheet. Advised for small budgets; big spenders can optimize with delayed events.

Daily spreadsheet: ad spend; new yearly free trials x 35 EUR projected value (replaced with actual conversions after 7 days); 50% offer revenue; new monthly revenue; total after store cut; P&L minus ad spend; and P&L minus ad spend minus a 900 EUR/day organic baseline (a deliberately high estimate of what social brings without ads). Above 0 means ads work. Needed because ATT consent is under 10% and SKAN data lags, and traffic comes from many sources (Instagram, affiliates, YouTube, Reddit, Facebook, ads). Uses RevenueCat filters per offer.

Judge a format by Instagram's 'watched past 3 seconds' share, plus average watch time (6-10s for his good formats) and engagement, not raw views. If a format consistently gets 75-80% past 3 seconds (all his good ones have 70%+), keep posting it: it will go viral eventually. On TikTok such a format gets at least 4-5K views per video. Consistent 4-5K is 'really good', and a viral hit is only a matter of time. Drop formats that get only double-digit views ('sub 50'). Expect luck: he once posted for a month at 81% with nothing going viral.

A manual Google/Excel sheet with one row per phone/account: account status and health, whether it's online, content type, notes, posts per day, plus daily views and likes; performance checked on day 2. A 1-hour team meeting every morning reviews it. Winners go back into the AI playbook; unhealthy accounts get recovery attempts, then are ditched and replaced. He avoids auto-tracking tools for own accounts because they make you stop looking at each piece (tools are fine for UGC at scale).