This is what we're building together over the next ninety days.

The mechanics that built ThirdHome to $20M will not get it to ten thousand SQLs. Pre-Andromeda audience targeting, monthly creative refresh against $127K of Meta spend, and Meta lead forms that physically cannot capture the SQL link-submission requirement are all artifacts of an earlier era. The fix is three pillars, executed in parallel.

The five clusters below were built by analyzing 813 substantial TrustPilot reviews — clustered on the language reviewers used to describe themselves, what they own, why they joined, and what they emphasized when recommending or not recommending the product. Persona size estimates are relative weights, not exact percentages. First-week deliverable is validating these against Carissa, the international team, and the sales team's lived experience.

A note on what this is and isn't. Eight hundred public reviews is a starting point, not a finish line. TrustPilot self-selects toward the enthusiastic and the angry — the much richer signal lives inside ThirdHome already: sales-call transcripts, member-services email threads, the twice-weekly feedback calls, churn conversations, the application data on prospects who didn't convert. Week one is access to that data and a refresh of these clusters against it. Persona size weights, channel-fit assumptions, and headline DNA all get sharpened or rewritten. The architecture below is the right shape; the numbers and the edges of each persona get tighter once the data behind the curtain is on the table.

The Unicorn assessment surfaced ten visual style buckets ThirdHome reacted strongly to. The library is organized as a matrix: each style runs against each persona, with four message angles per cell. That's the production target by month three. Eight wave-one ads are produced and visible below — four for the Empty-Nest persona, four for Multi-Gen, balanced 50/50 between luxury polish and pattern interrupt. AI production is the only way to ship at this volume against $127K/month Meta spend.

Wealthy second-home owners take 60 to 90 days to commit to anything. They don't fill out a form because of one ad. They fill out a form because the third ad answered an objection the first one raised, the fifth ad showed a member who looked like them, the seventh ad happened to land in a week when their property tax bill arrived, and the eighth ad was the math finally cleaner than they could ignore. The current Meta account treats every impression like it's the conversion impression. It isn't. It's a touchpoint inside a sequence — and the sequence is what gets us to ten thousand SQLs.

Full-funnel messaging maps every documented objection to a content asset, repurposes that asset into ad creative, and stacks the creative into time-and-engagement-based retargeting audiences. A prospect who first saw a pain-hook ad on day zero sees the objection-handler content on day seven, the math on day twenty-one, the member voice on day forty-five, and the soft "schedule a call" on day sixty. Each touch is the next sentence in a conversation, not a fresh first impression. This is what direct response gets wrong about luxury buyers — and what fixes the math.

Eight specific objections every slow buyer raises before they SQL. Each one becomes a content asset, each content asset becomes ad creative. One piece of content = four to six ads. This is the structural fix to "creative refresh too slow" — the math finally works at $127K/month spend because we're producing creative variations, not new concepts every week.

Five audiences. Each defined by engagement depth and time since first touch. Each shown a creative bundle calibrated to that point in the consideration cycle. The prospect doesn't see the same ad twice in a row — they see the next ad in the conversation. Frequency caps prevent fatigue. Audience exclusions prevent overlap.

Two personas, three formats each — sales page, advertorial, listicle. Each format hits the same persona from a different angle: the sales page does pain-to-solution, the advertorial does first-person narrative, the listicle does educational reframe. Six pages give us six creative directions to A/B against each persona's audience, isolating which structure converts best for whom.

Each page is built in ThirdHome's existing brand register, with the SQL link-submission qualification gate built in. Each routes to the same SQL form. Each fires a custom conversion event tagged with persona × format attribution so Andromeda learns which combination converts best.

A note on Persona 03 (Branded-Residence). Branded-residence members come through partnership channels — comp'd trials inside Hyatt VC, Vidanta, Marriott Vacation Club, Grand Solmar. That funnel needs in-resort placements, partner emails, and trial-conversion sequences, not paid-social LPs. ThirdHome's stated paid-acquisition priority is whole-home owners. Persona 03 is documented in the persona architecture above for internal channel intelligence and future partnership-channel investment, but is intentionally not in this paid-acquisition LP set.

The current US account is fragmented into multiple narrow-targeted ad sets that pre-date Andromeda. The restructure consolidates into three persona-themed campaigns, a creative testing lab, and a retargeting campaign — total of five. Each persona campaign runs 15–20 ads on broad targeting, mixed format, optimized for SQL with link submission. The math gets cleaner. The signal back to Andromeda gets stronger. The CPA drops.

EMEA tier-2 geographies (Italy, France, Spain, Germany) get separate ad accounts during the international takeover in month five. Until then, EMEA stays in the existing structure to avoid disrupting in-flight campaigns. The audit recommendation for separation is correct; it gets executed once US trust is established.

The +30% YoY SQL growth mandate is the number that defines the engagement. Against the 2024 baseline of 7,281 SQLs (the most recent confirmed annual figure in the Unicorn brief), that's roughly 2,200 incremental SQLs annually. With Meta carrying about half the load on a typical year, that's an additional ~90 SQLs per month from Meta against current run-rate. The persona-first architecture exists to make those incremental SQLs available without proportional spend increases — and validating the actual current-year baseline against ThirdHome's internal Looker data is week-one work, since the public brief is a year stale.

Day-thirty deliverable: a measurable bend in the SQL trend that's been declining over the last forty days. Day-ninety deliverable: a documented win/loss creative library, a clean Andromeda-native US account, and a scaling brief for months 4–12.

Eight specific actions that ship inside week one. Half are stop-the-bleed (the audit findings nobody pulled). Half are foundation moves that compound over months — the SQL list seeded as lookalikes, the link-submission custom conversion, the creative-tracking spreadsheet that fixes the recurring "we lose track of what worked" frustration.

The Unicorn assessment surfaced specific frustrations from past agency relationships. Each one has a structural fix in this engagement. None of them are addressed by saying "we'll do better." They're addressed by changing the operating model.

The 813 TrustPilot reviews that informed the persona work are the public-facing tip of a much bigger data set. The much more valuable insight lives inside ThirdHome already — sales call transcripts, member-services email threads, the twice-weekly feedback calls, and the cancellation conversations. With access to that, the creative pipeline becomes self-feeding. This is the moat no agency without internal data access can replicate.

I don't replace anyone on this list. I sit between them and make the system run faster. Damien stays through the transition. Carissa keeps building. The landing page consultant keeps owning paid search LPs. The sales team keeps the lead-quality signal. Nothing here disrupts existing relationships — every one of them gets stronger because the pipeline that flows through them is sharper.

I do not need answers to these before week one. I need answers to these during week one. Each of these shapes the priorities of the first thirty days. Each of these is an invitation to align before the engagement gets expensive to redirect.