Hitting the call target is not a spending problem — it is a throughput and creative-supply problem. The realistic relaunch target is ~17 booked calls a day on two closers for the first ten days, climbing toward a 28-a-day north-star once the third closer lands — because closers close five days a week but ads run all seven, so the right divisor is 7, not 5. At $100 or more per qualified call, 28 a day is about $3,000–$4,000 a day in ad spend. We have 7 proven winners to switch straight back on, and — reviewed 15 June — the full 26-ad standby reserve is now genuinely ready (the ads that had been sitting at first-pass review are signed off). With the winners, that is ~33 ads ready to run on day one, plus 15 finished AI-hook variations ready to layer in, with 3 more (PassionToProfit) being spliced now. What limits us next is throughput — producing genuinely new angles before the reserve runs down — plus lifting the show-up rate, which adds held calls without spending a dollar more.
The relaunch brief asked five things. Here they are at a glance — detail follows below.
| The question | The short answer |
|---|---|
| How many ads are live and ready to turn on? | 30 ran last cycle — 7 are proven winners we switch straight back on. All 26 standby ads are now reviewed and ready (signed off 15 June), so with the winners that's ~33 ready day-one, plus 15 finished AI-hook variations ready to layer in (3 more PassionToProfit hooks being spliced now). All Amelia. |
| How did they perform? | Winners booked calls at $25–$72 each (target is $150). Blended across the whole run, about $79 a booked call — last run, before tightening qualification. Strong base. |
| What changes to hit the target? | A staged spend ramp: switch winners on → scale daily → consolidate. ~$1.5k/day → $3–4k/day. ~17 calls/day on 2 closers (first 10 days) → ~28/day once the 3rd closer is in. |
| How much is genuinely ready, by angle and format? | 26 standby, all reviewed and ready — proven angles, about a third in fresh formats — plus 15 finished AI-hook variations (3 more being spliced). A comfortable few weeks of runway. The 27 "in production" are mostly parked in engineering — real future supply, not near-term. |
| When do we need to produce more? | With the 26 standby now cleared and the lower call target, a comfortable few weeks. Then 5–10 genuinely new ads a week — the one real gap is new angles, which internal production cannot sustain alone. Decision needed |
30 ads went live in the March run. Performance splits cleanly into three groups — and a small handful of winners did almost all the work.
The seven winners and the angle each speaks to:
| Ad | Angle | Format | Cost / booked call | Booked calls* |
|---|---|---|---|---|
| Design Revolution | Credentials | Selfie | $25 | 16 |
| Student Results | Market | Studio | $54 | 32 |
| Pinterest Profit | Market | Studio | $58 | 8 |
| Passion to Profit | Market | Selfie | $63 | 26 |
| No-Degree Success | Credentials | Studio | $64 | 66 |
| Accountant Escape | Escape | Selfie | $70 | 17 |
| No-Degree Success | Credentials | Selfie | $72 | 31 |
* Booked calls = Hyros attribution across the last run (Mar–Apr 2026). Cost per booked call from the ad-platform reconciliation.
The shape is unmistakable: of the 264 booked calls across the run, the seven proven winners drove 196 — 74%, led by No-Degree Success at 97. Credentials and Market are the dominant angles, with Escape close behind. A few concepts always carry the account — this tells us exactly where to point spend on day one, and which proven angles to re-stock first.
One more to watch: the Speed angle (Jenny / First Week) booked 16 calls — real volume, but at a cost per call well above the $25–$72 winner range. Re-test it as a fresh angle, not a switch-on-day-one winner.
Reconciled against the editing pipeline and reviewed on 15 June. All are confirmed Amelia ad creative; none are in Ads Manager yet. Here is where the reserve genuinely stands.
| Bucket | Count | What it is |
|---|---|---|
| Ready to publish | 26 | Edited, reviewed and signed off (15 June) — the day-one reserve beyond the 7 winners. Proven angles, about a third in fresh formats. |
| AI-hook variations | 15 ready +3 in splice | Fresh AI-generated hooks on the 5 winners, split into their own creatives — each a new Entity ID. 15 finished and ready to layer in; 3 PassionToProfit hooks being spliced onto the body now. |
| In production (the "27") | 27 | Filmed or engineered, mostly parked in engineering — real future supply, but it needs an owner to finish. Not near-term runway. |
| Scripts only | 19 | Written, not filmed — including 3 unique new angles just sent to engineering. |
Day-one inventory is real: 26 reviewed standby ads + 7 proven winners ≈ 33 ready to run, with 15 finished AI-hook variations ready to layer in behind them, and 3 more PassionToProfit hooks being spliced now (bringing the total to ~48 once complete). The "27 in production" is still not near-term — mostly parked in engineering, it needs an owner to finish. So the reserve comfortably covers the first weeks; the one real gap is genuinely new angles (below) — which is why 3 unique filmed concepts were just pushed into engineering, and new-angle production needs to spin up now.
The standby set by angle — every one is a proven angle. There are no brand-new angles in Amelia's reserve:
| Angle | In standby |
|---|---|
| Market | 6 |
| Credentials | 4 |
| Escape · Identity · Flexibility | 3 each |
| Overwhelm · Speed · Objections | 2 each |
| External | 1 |
By format — mostly the proven studio look, about a third in fresh formats:
| Format | In standby | Role |
|---|---|---|
| Studio (direct-to-camera) | 18 | Proven format — re-stock the winning angles |
| FAQ | 3 | Fresh format — new audience pockets |
| This-or-That | 3 | Fresh format |
| Seated Interior | 1 | Fresh format |
| Podcast | 1 | Fresh format |
The ad algorithm now decides who sees each ad based on how different it looks and sounds. Near-identical ads get merged and starved of reach, so genuinely new angles are how we reach new pockets of buyers and scale without burning the winners out. Amelia's reserve gives us solid re-stock of the proven angles and a handful of fresh formats — enough to relaunch and sustain the first weeks. But it holds no new angles. The brand-new angles in the database belong to the other experts, not Amelia. So new-angle production for Amelia is the thing to spin up now, not later.
From a paused account you cannot land the number overnight — the algorithm needs a few days to re-learn. So we step spend up deliberately, on two closers through the first ten days, watching cost per call hold before each increase, then push toward the 28-a-day north-star once the third closer is in.
| Window | Daily spend | Booked calls/day | The move |
|---|---|---|---|
| Days 1–3 · re-stabilise | ~$1.5k | building | Switch the 7 winners back on (same ads, so they keep their existing likes and comments) + the first push-ready standby. Let the algorithm re-learn. |
| Days 4–10 · climb | ~$2–2.5k | ~13–17 | Scale budget up in steps once cost per call holds under ~$100–150. Layer standby into the winners to widen reach. (2 closers full-time — the safety buffer.) |
| After day 10 · sustain | ~$2.5–4k | ~17 → 28 | Consolidate the proven winners into one optimised campaign and push spend. Layer in the fresh-angle standby. (3rd closer confirmed → climb toward 28.) |
Turn on → scale up → diversify → consolidate. We do not rebuild anything on day one — the winning ads are paused, not gone, so we simply switch them back on and raise the budget. We only ever raise budget by 10–20% at a time and watch the trend, so we never spook the algorithm. "Duplicating" ads only happens at the very end, once we gather the winners into one campaign to scale cleanly.
Good closers are the slowest variable to hire and the easiest to lose, so we start with two full-time for the first ten days — every call reviewed personally — and bring the third on only once the sales process is bedded in. Marketing deliberately runs ahead of sales capacity (the 28-a-day north-star is a >110% ceiling, not a forecast) so creative and engineering are never the thing holding growth back.
We deliberately aim the calendar slightly above target so a soft day on show-up rate never leaves a closer with empty slots.
Booked calls only matter if people show up. Here is the chain from ad spend to a closer's calendar, for both the two-closer start and the three-closer north-star — and the single cheapest way to add more held calls.
| Step | 2 closers · first 10 days | 3 closers · north-star |
|---|---|---|
| Booked calls a day | ~17 | ~28 |
| Show-up rate | 40–50% Provisional | 40–50% |
| Held calls a day | ~8–9 | ~13–14 |
| Per month | ~510 booked → ~250 held | ~840 booked → ~400 held |
Why the divisor is 7, not 5: closers take calls five days a week, but ads run all seven and the weekend's bookings fill the week ahead. 28 a day is the >110%-capacity ceiling we steer toward; the realistic figure lands lower until the setting process and show-up rate are dialled in.
A 40% show-up rate is normal for this kind of funnel — the industry sits at 30–40%, so nothing is broken. But it is also the cheapest thing to improve. Lift it from 40% to 55% and the same 28 booked calls go from ~11 held to ~15 held a day — roughly a third more held calls, for zero extra ad spend. The levers are known: a simpler, clearer confirmation page, a one-click confirmation, a short personal message before the call, and framing the call as the most important meeting of their week. A well-oiled setting process (the work Valentino is building) is the unlock. This deserves its own focused push alongside the ramp.
The reserve buys us time, not forever. Here is the honest read on how long we are covered and what it takes to stay ahead of ad fatigue.
A lower call target (~17→28/day, not 40) and a reviewed, ready reserve give a comfortable few weeks of runway: ~33 ready ads with the winners carrying most of the volume, plus 15 finished AI-hook variations to layer in as the originals fatigue (3 more PassionToProfit hooks being spliced now — ~48 total once complete). The "27 in production" and the 19 scripts are real future supply on top of that — but only once someone owns finishing them. The clock that matters is not this fortnight; it is new-angle production, which has to be spinning up now so it lands before the proven-angle reserve burns down.
Only 1 in 6 new ad concepts wins (15–18% — this is the rate top agencies see, and it matches our own run). So producing new ads is an investment where most "fail" by design. To keep finding winners as old ones fatigue, we need volume: 5–10 genuinely different new ads a week, which yields 5–6 new winners a month. That is enough to replace fatigue and keep growing. They must be genuinely different (new angle, face, or format) — minor tweaks get ignored by the algorithm.
The bottleneck is throughput, not ideas — and with the review queue now cleared, two stages remain. First, engineering ownership — the 27 "in production" (and the 3 unique new angles just queued) need someone accountable for finishing them, or they stay parked. Second, the durable one: based on the last 15 months, internal production alone will not hold 5–10 genuinely-new ads a week once the reserve is gone, and Amelia's reserve is proven-angle re-stock with no new angles in it. The two ways through: (a) bring on real people to film as new faces — every new face is guaranteed fresh reach — and dedicated editors, or (b) fund full AI-generated ad hooks at volume. Most likely both. It needs a decision and hiring time now, not when the reserve runs dry.
Producing 5–10 net-new ads a week is the long game. The faster, cheaper lever — especially below ~$3k/day — is to squeeze more life out of the winners we already have, and to spread production deliberately so the algorithm keeps finding fresh audiences.
A winning ad has a finite half-life before fatigue sets in. The cheapest way to extend it is to re-cut the first three seconds — a new hook on the proven body — rather than build a new ad from scratch. One winner becomes three or four variations, each a fresh shot at a new pocket of buyers, at a fraction of the production cost. Our creative database already tracks this as a next step on every ad (iterate with new hook · re-cut body), so it is a workflow we can run, not a guess.
Meta now rewards distinct-looking creative with reach into new audience pockets — each materially different ad earns its own "Entity ID" and a fresh range of spend, while near-identical ads get merged and starved. So production should be allocated on purpose across angles, faces, and formats, not piled onto one look. Iterating winners and diversifying deliberately is how we unlock more spend before we have a flood of brand-new concepts — and it is the cheapest needle-mover at low spend. Scoring framework: Winner Iteration
A sense of the prize at our proven efficiency — a forward projection of what ~28 booked calls a day (the three-closer north-star) produces.
| Line | Figure | Note |
|---|---|---|
| Ad spend to run ~28 calls/day | ~$3–4k/day (~$90–120k/mo) | Assumes $100–150 per qualified call after tightening qualification — above last run's ~$79 |
| Held calls a month | ~400 | ~840 booked at ~50% show-up |
| Sales a month (ramp) | ~50–60 | At an early 12–15% close rate (month 1), rising to 15–20% (month 2) and ~25% (month 3) |
| Program value | ~$6,000 | Contracted value per student |
| Cash collected upfront | ~$597/sale today | The gap to close — payment-plan deposits, not full value |
The distance between the ~$6,000 program value and the ~$597 collected upfront is the biggest lever on whether the ramp is cash-flow positive from day one. Cash collected is still minimal, so we do not yet model cash or revenue ROAS in any meaningful way — the metric we build toward is CAC : LTGP (cost to acquire a customer vs the lifetime gross profit they bring), the truest read on whether a customer is worth more than they cost to win. Improving deposit and payment terms may move the picture more than any ad optimisation. First step is to centralise this data.
A handful of calls to make. The first ones are time-sensitive — they need action now, not when the reserve runs dry.
| Decision | Why it matters | When |
|---|---|---|
| Give the "27 in production" an owner to finish (or shelve) | With Fibery now reconciled and the dead records cleared, the 27 is the last unmanaged supply — it needs someone accountable or it stays parked. | This week |
| Creative production past the reserve — hire filming faces + editors, fund AI hooks, or both | The true constraint on sustaining the climb. Needs hiring lead time now. | This week |
| Stand up a focused show-up-rate push (confirmation page + pre-call nurture + setting process) | The cheapest held calls we can add — free capacity | This week |
| Confirm timing of the 3rd closer | Start with 2 full-time for 10 days (deliberate), then add the 3rd to climb from ~17 toward ~28/day | ~Day 10 |
| Centralise sales + cash-collected data | Unlocks the real profit picture and the payment-terms lever | Week 1 |
Sign off on the staged ramp shape (turn winners on Monday/Tuesday → ~$1.5–2.5k/day to ~17 calls on two closers through the first ten days → toward ~$3–4k/day and ~28 calls once the third closer is in), and greenlight the creative-production decision so the calendars stay full when the reserve runs down. Day-one inventory is ready — ~33 ads plus 15 finished AI-hook variations (3 more PassionToProfit hooks being spliced, ~48 total once done) — so the live call now is standing up new-angle production before the proven-angle reserve burns down. Everything else in this report is detail in service of those moves.
| Booked call | A prospect who scheduled a sales call from an ad. |
| Held call (show-up) | A booked call where the prospect actually turned up. |
| Cost per booked call | Ad spend divided by booked calls. Our winners ran $25–$72; our target is $150 or less. |
| Angle | The core idea an ad speaks to — e.g. "you don't need a degree" (Credentials) or "escape your current career" (Escape). |
| Format / style | How the ad is shot — studio direct-to-camera, selfie, FAQ, podcast, and so on. |
| Standby | A finished ad, ready to publish, not yet turned on. (All 26 reviewed and genuinely ready as of 15 June.) |
| North-star target | A deliberate above-capacity ceiling we steer toward — here, 28 calls/day on 3 closers — not a forecast. It keeps marketing ahead of the slowest-to-hire function (closers). |
| Winner iteration (half-life) | Re-cutting a proven ad's hook (or body) to extend how long it performs before fatigue, instead of building a new ad from scratch. |
| Entity ID | Meta's signal that an ad is materially different from the rest — distinct creative earns its own reach and a fresh range of spend. |
| Creative fatigue | When an ad has been seen too often and stops performing — fixed by feeding in fresh ads. |
| Close rate | The share of held calls that become paying customers. |
| Qualified call | A booked call from a prospect who meets our fit criteria. Tightening qualification means fewer but better-matched calls — at a higher cost each. |
| ROAS | Return on ad spend — revenue earned for every dollar spent on ads. |
| CAC : LTGP | Customer acquisition cost vs lifetime gross profit — what it costs to win a customer against the profit they bring over time. The ultimate measure of whether the maths works. |