Find the local business stuck behind. Build them ahead — before we ever call. Charge every month. This document maps the whole machine: what it costs, what we charge, who does what, and the numbers that get it to $200k → $500k recurring.
Every web agency sells a promise. Builtfore sells a finished product. Software enumerates every business in a suburb, scores how bad their web presence is, and auto-builds each top prospect a personalised homepage on a live preview URL — theirbusiness.builtfore.com.au — before the first phone call. The pitch is one sentence: "I already built it. It's yours."
software gross margin — prospecting costs cents, hosting is near-free
rent it (subscription, the asset) or buy it (cash up front, the easy yes)
the only cost that scales is people — the model runs on one seat
Australia has ~2.73 million actively trading businesses (ABS, June 2025); 97% are small. The serviceable slice — local service businesses found via Google search — is roughly 700k–900k. Estimates of how many have no website range from ~22% to 59% depending on the study; even the conservative end, plus the large share sitting on dated, slow, non-mobile sites, puts the addressable pool at 300k–500k businesses. Consumer pressure is real: ~75% of consumers will only buy from a business with a website; 40% won't buy from a social-media-only presence.
Plumbers + electricians + mechanics across ~15 Western Sydney suburbs, mid-range assumptions:
Places API lists every business in a category + area: name, phone, website, rating, review count. Fully automated.
Fetch each site; check platform fingerprints (Wix / GoDaddy / Squarespace), mobile viewport, HTTPS, PageSpeed. No-website businesses flag as top prospects. Fully automated.
Badness × (rating × review count). Surfaces the 4.8★, 200-review business on a broken 2015 site; buries the 3★ with 6 reviews. Fully automated.
Pull logo, photos, real reviews, hours, services → premium template → live preview URL in minutes. ~90% automated; every mockup gets a 3–5 min human QA before it's ever sent (wrong logo or bad photo burns the pitch).
Call, send the live link. The puppy-dog close — no imagination risk. This stage is human and always will be; automating it (mass email) converts dramatically worse.
| Path 1 · Subscription (the asset) | Price | Includes |
|---|---|---|
| Starter — "Get found" | $89/mo | 5 pages · mobile · SSL · hosting · lead form · minor monthly edits |
| Growth | $199/mo | 10 pages + Google Business optimisation + review automation + booking |
| Authority | $449/mo | Everything + SEO content engine (~4 posts/mo) + monthly report + priority |
| Add-ons | $25–349/mo | SEO package from $349 · extra pages $25 ea · CRM/booking/SMS module $49–99 |
| Path 2 · Buyout (the cash) | Price | Includes |
|---|---|---|
| Build & handover | $1.9–2.9k | ~5 pages, handed over on their own hosting — they own it outright |
| Care plan (the hook) | $49/mo | Updates, hosting help, support — converts a buyout into recurring |
| Optional setup fee | $199–499 | On subscriptions — offsets caller commission |
Recurring revenue only — buyouts are cash on top. The number that moves everything is blended ARPU: what the average client pays once add-ons are counted.
| Blended ARPU / mo | → $200k | → $300k | → $500k |
|---|---|---|---|
| $120 · mostly Starter | 139 clients | 208 | 347 |
| $170 · healthy mix | 98 | 147 | 245 |
| $250 · good SEO attach | 67 | 100 | 167 |
| $350 · strong attach | 48 | 71 | 119 |
Planning range: 8–15 net new subscribers/month once the script is proven (the original 15–25 figure is the ceiling, not the base case). A couple of $2k buyouts a month covers the seat — the subscription base compounds on the deals' own cash. Second seat = the $500k lane, once the script is proven.
| Component | Rate | Why it's shaped this way |
|---|---|---|
| Buyout builds | 20–25% of build | $400–700 per close, paid on collection — the caller's fast money |
| Subscription bounty | 1× first month | $89–449 per close, immediate reward per sub |
| Subscription trail | 10% × 12 mo | Trails stack — same performance pays more every month; the retention hook |
| Add-on attach | 15% trailing | Aligns the caller with ARPU — the real lever — not raw count |
| 90-day clawback | Bounty netted | Churn inside 90 days costs the caller too — stops bad-fit closes |
Illustrative: 8 subs + 2 buyouts/mo ≈ $1,800–2,500 in month one, rising to $3,000–4,000/mo by month six as trails stack — the compounding annuity is what keeps a commission-only seat loyal.
The founder's work is asynchronous and batchable; only the calling is time-of-day dependent. The two roles never need to be online together.
Run enumeration + scoring for the next region (~30 min babysitting) → review ranked list, kill duds (~45 min) → generate + QA 20–30 mockups (~90 min) → load tracker with phone, link, badness reasons, suggested opener (~20 min).
Works the list top-down. Letters → calls → SMS link → logs every outcome → books closes. Owns all client-facing conversation through a shared inbox. Founder never needs to be present.
One 20–30 min check-in (or voice notes): what's landing, what objections repeat. Lagged answers to caller questions. Script + scoring tweaks from what the calls reveal.
| Phase | Founder time | Where it actually goes |
|---|---|---|
| Months 1–2 | 10–15 hrs/wk | Scripts, pitch iteration, first fulfillments — and listening to early call recordings |
| Steady state | 8–12 hrs/wk | Mostly fulfillment: each closed build costs 4–8 hrs even with the template |
| 30+ clients | Trap zone | Support creep stops being batchable — systematise edits or hire before this point |
Before scaling anything: one region, one category set, 50 scored prospects, mockups for the top 20, and 100 real calls over four weeks. This replaces every modelled number above with a measured one.
2+ buyouts or 4+ subscriptions close in the first 100 calls. The machine works — add regions, then the second seat.
Zero closes means the script or the market slice is wrong — not necessarily the model. Fix the pitch on real objections before spending another dollar.