Attribution & offline conversions — Healthcare & high-ticket
Your ad platform doesn’t know which clicks became patients. Mine does.
I’m Lev Brovtsev. I rebuild the measurement layer between your ad accounts and your CRM: the click ID is captured on every lead, and when that lead becomes a verified patient or a closed deal, the outcome is imported back into Google Ads and Microsoft Ads. Bidding then optimizes toward money you have actually collected — not toward form fills a platform decided to count.
You work with me directly, one-on-one. No account managers, no juniors, no handoff. Four clients at a time.
15 minutes. With me, not a sales rep.
The problem
Two reports. Two realities. One of them is spending your money.
Your agency’s dashboard says 140 conversions last month. Your intake team remembers eleven consultations and four patients. Both numbers are “true” — they’re just measuring different things, and the ad platform is only allowed to see the flattering one.
So the algorithm optimizes for what it can see. It finds the campaigns, keywords and placements that produce the cheapest form fills, and it pours budget into them. If those form fills happen to be tire-kickers, price shoppers, or the wrong procedure entirely, nothing in the system objects. Cost per lead falls every month while cost per patient quietly climbs.
This is not an agency being lazy. It’s a measurement gap. Until the outcome that matters travels back to the platform, every bid you place is a guess dressed up as automation.
| Source | What gets counted by default | What it misses without closing the loop |
|---|---|---|
| Web form | Every submission, spam included | Whether the submitter was ever contacted, let alone treated |
| Phone call | Calls past a duration threshold | Wrong numbers, vendors and support calls clear that bar too |
| Chat / booking widget | Widget opened or started | Whether the conversation ever produced a real appointment |
| Walk-in / referral | Usually not tracked at all | The channel that originally built awareness gets zero credit |
What I actually do
Close the loop — click to CRM and back again.
GCLID and MSCLKID on every lead record
Click IDs are captured at the landing page, carried through forms, chat and call tracking, and written to a dedicated field on the lead in your CRM. If the ID is lost at intake, nothing downstream can be recovered — this step is where most implementations quietly fail.
Pick the stage where money is real
Consultation booked, consultation attended, treatment accepted, closed-won. We choose the earliest stage that reliably predicts revenue and still gives bidding enough volume to learn from — usually not the stage your agency has been reporting.
Offline conversion import into Google and Microsoft
Outcomes flow back on a schedule — from Salesforce through Google's Data Manager, or via the Data Manager API and scheduled uploads for HubSpot, Zoho and custom systems. Conversion values carry real revenue, so value-based bidding has something honest to maximize.
PHI stays on your side of the line
Only a click ID, a timestamp, a conversion name and a value leave your systems. No names, no contact details, no clinical information — a materially safer architecture than the pixel setups most clinics are running today.
Rebuild bidding on the verified signal
Target CPA and target ROAS are reset against verified outcomes, budgets move toward campaigns that produce patients, and the ones that produced cheap noise get cut. This is the point where reported performance gets worse and the business gets better.
One number the owner can act on
Cost per verified patient, by channel, by campaign, reconciled to the CRM. If a report can’t be traced back to a record in your system, it doesn’t go in the report.
Proof
What happens when the signal gets honest.
A luxury US healthcare clinic. I inherited a wasteful account, rebuilt it around verified patients imported from the CRM, then scaled. Relative figures only.
"He closely monitors the full lead generation flow across different systems and CRMs, making sure data is accurate, attribution is clear, and nothing falls through the cracks. Lev consistently follows each lead through the pipeline, checks tracking integrity, and works across platforms to ensure everything is aligned and performing as expected."
"Lev immediately stands out as an exceptional partner to work with over my last 3 years with Google. He uniquely blends high-level, data-driven strategy with deep, hands-on technical skill — precisely the kind of results-oriented partner you want managing significant in-platform investment."
How we work
Prove the measurement first. Everything else follows.
$500 paid audit, credited toward setup
I map what your account currently counts as a conversion, where the click ID is being dropped, and what your CRM would have to expose to close the loop. You get the roadmap whether or not we work together.
Implementation, two to four weeks
Tracking, CRM fields, call tracking, the import schedule, and the conversion actions in both ad platforms. Done once, properly, with your developer or mine.
Monthly retainer, direct with the owner
Ongoing management across Google, Microsoft and Meta as one system, priced against verified patients. Prepaid monthly, no account managers in between.
Questions owners ask
Before you book a call.
What is offline conversion tracking?
Offline conversion tracking sends outcomes that happen after the click — a consultation attended, a treatment accepted, a deal closed — back into the ad platform. The click ID (GCLID on Google, MSCLKID on Microsoft) is stored on the lead record in your CRM, and when the record reaches a revenue stage that outcome is imported back against the original click. The platform then optimizes toward the outcome you actually sell. None of it works unless each lead first carries its source; that groundwork is lead source tracking.
Do I need Salesforce for this?
No. Salesforce connects directly through Google's Data Manager, which makes it the fastest path, but the same loop works with HubSpot, Zoho or a custom CRM through scheduled uploads or the Data Manager API. What matters is that your CRM can store a click ID against the lead and expose the stage at which money is confirmed.
Is this HIPAA-safe for a clinic?
It can be, and it is safer than most pixel setups. Nothing about a patient leaves your systems: what is sent back to the ad platform is a click ID, a timestamp, a conversion name, and optionally a value. No name, no email, no condition, no PHI. The design point is to keep patient data on your side of the line while still letting bidding learn.
How long before bidding improves?
Implementation usually takes two to four weeks depending on how your CRM stages are set up. Smart Bidding then needs enough imported outcomes to retrain, which in high-ticket healthcare typically means another 30 to 60 days. It is slower than switching a bid strategy — and it is the only change that makes the bid strategy honest.
My agency says conversions are up. Why does this contradict them?
Because a platform conversion is a form fill or a call connect, and a modeled conversion is a statistical estimate of one. Neither is a patient. When the CRM outcome is imported, you usually find that a large share of reported conversions never reach a revenue stage — and that the campaigns producing the most conversions are not the campaigns producing the most patients.
Reference: Google Ads offline conversions · Microsoft Ads offline conversions · HHS guidance on online tracking technologies
Built on Google's documented methods: offline conversion imports and enhanced conversions for leads.
Get started
Let’s find out what your ads are actually producing.
A 15-minute Fit Call to confirm whether we fit and agree the next step. Not an audit, not a free consultation; the diagnosis happens in the paid audit.
Prefer to start small? A $500 paid audit, credited toward setup is the fastest way in.
You work with me directly, one-on-one. No account managers, no juniors, no handoff. Four clients at a time.
Book a Fit Call15 minutes. With me, not a sales rep.