leads in a single 30-day window, every one tracked to its source.
The proof, in full.
One practice, one month, every number and where it came from, including what we don't claim.
A cash-pay urology subspecialty practice in West LA.
The challenge
A cash-pay urology subspecialty in West LA, focused on men's sexual health and fertility. The patients are sophisticated and private. They research a condition for weeks before they ever call, and they choose on reputation and outcomes, not on which office is in-network.
That makes the usual medical-marketing playbook useless. Insurance-volume tactics, Medicare panels, and review farming don't reach a patient quietly comparing subspecialists at midnight. The practice needed trust built before the first call, and a system that could capture and route demand by condition.
The system
We installed the Compound: an in-house video pipeline for condition education, paid ads on Meta with every lead tracked to the specific ad, and an automation layer on GoHighLevel (run on its HIPAA-compliant tier) with an AI bot answering first questions in DMs. Every lead is tagged by condition and routed to its own follow-up sequence.
Underneath it all runs the tracking stack: Google Analytics 4, CallRail, and the Meta Pixel, pulled into one view through Windsor.ai, so every form and call ties back to the source that produced it. The practice watches it on a live dashboard and gets a written report every month. The full methodology is here
The results
Over nine months the content engine built the reach. By month nine, that reach was converting into about 111 leads every 30 days across 11 distinct sources, roughly half by form and half by tracked phone call.
Figures from CallRail and form-submission exports (30-day window, 2026) and a Metricool report, for a cash-pay urology subspecialty practice in West LA, anonymized for client privacy. These are attention, traffic, and lead figures, not booked appointments, and not a claim about patient volume or revenue.
The insight
Organic outproduced paid by almost two to one. The work shipped in month one is still producing leads in month nine. That's the long play, proven.
ChatGPT and SearchGPT are already showing up as referral sources. Practices using the Compound are already cited by the search engines patients will use in three years.
Three confirmed AI-search leads in the last 30-day window. Small today, but the trajectory is the point.
Curious what the system would do for your practice?
What we count, and what we don't claim.
These are leads, form submissions and tracked phone calls, not booked appointments. We don't put a patient count or a revenue figure on this page, because we can't tie a specific lead to a specific patient with permission to share it, and we won't imply data we can't stand behind. Of the roughly 111, about 93 were substantive. We filter out calls under a minute rather than pad the number. Every figure here comes straight from the attribution data, reconciled across channels, and held to a standard we'd defend on a sales call.
Roughly half of this practice's leads arrive by phone, not by form. We track both with CallRail and reconcile them into one view, so phone calls count as the leads they are.
Search, social, paid, direct, referral, and AI-search referrals all land in one picture instead of five disconnected dashboards.
GA4, the Meta Pixel, and our tagging trace each lead back to the specific ad, post, or page that produced it. No guessing which channel worked.
Public numbers are defensible and anonymized. We protect client identity until a client approves being named, then swap in the name and add specifics.
Your practice could be the next case here.
We work with one cash-pay practice per subspecialty per metro market. Cash-pay urology in the LA market is committed. Plastic surgery, psychiatry, and podcaster slots are open in most markets, along with personal-injury law firms. If your market is open, the system is ready to install.
Ready to build the system?
30-minute strategy call. No pitch. We map your current setup and tell you what we'd build.