Every month, thousands of businesses sit in reporting meetings celebrating a number that means almost nothing: total leads. It's the metric everyone reports because it's the metric that's easy to count — and it's routinely inflated by spam form fills, SEO solicitations ("I visited your website and noticed..."), job seekers, existing customers calling back, and inquiries for things you don't even sell.
We stopped counting that number for clients. This is the system we run instead, with the real results it produced for Prosper Plastic Surgery: verified good leads up 38% in five months, 347 total, on a budget that didn't grow by a dollar.
What counts as a 'verified good lead' (and what gets thrown out)
Every inquiry gets individually scored. Phone leads are rated on engagement — answered vs. missed, call duration, first-time-caller status. Form leads are scored on procedure specificity, decision stage, message depth, financing interest, and contact completeness. A lead only counts if it rates Strong or Good and it's for a service the business actually offers.
Excluded from every number we report: spam and solicitation submissions, fake contact details, repeat callers who are existing patients, off-target requests, and weak or abandoned contacts. What remains is what the business owner actually means when they say "lead": a real person, with real intent, for the right service.
If your lead count includes spam and job seekers, you're not measuring marketing. You're measuring form submissions.
Why this isn't just reporting hygiene — it changes the machine
Here's the part most people miss: verified scoring isn't a cleaner scoreboard, it's a better steering wheel. We import lead outcomes back into Google Ads (closed-loop reporting), so the campaigns optimize toward inquiries that become real consultations instead of toward whatever fills a form fastest. The algorithm learns what a *good* lead looks like — which searches, which ZIP codes, which landing pages — and shifts budget accordingly.
That's how the number moved without new spend: full-funnel tracking with CallRail, closed-loop outcomes into Google Ads, procedure-specific landing pages, precision geo-targeting, and AI-driven negative keyword expansion cutting the garbage before it ever generated a click. The full case study has the month-by-month.
Prosper Plastic Surgery, January–May 2026
- verified good leads/month (58 → 80)
- +38%
- verified good leads/month (58 → 80)
- verified good leads in five months
- 347
- verified good leads in five months
- added to the monthly ad budget
- $0
- added to the monthly ad budget
How to start scoring your own leads this month
- Instrument everything first. Every call and form through one tracking layer (we use CallRail) with source, page, duration, and first-time status. Untracked inquiries don't exist.
- Write your scoring rubric. What makes a lead Good for *your* business? Service match, decision stage, contact completeness. Keep it to one page so it actually gets used.
- Score a full month before changing anything. Your real good-lead rate is the baseline every optimization gets judged against — and it's usually a sobering number.
- Close the loop. Feed outcomes back into your ad platforms so the machine optimizes toward quality. This is the step that turns reporting into performance.
- Report the verified number to leadership. Smaller than the old number, and the first one that correlates with revenue. That trade is the whole point.
Quick answers
Won't my reported lead count drop when we start filtering?
Yes — commonly by a third or more — and that conversation is uncomfortable exactly once. Frame it as a currency conversion: the old number was inflated units, the new number is real money. Every decision made on the verified number (budget shifts, landing page calls, channel cuts) gets smarter immediately, and within a quarter nobody wants the old metric back. The businesses that refuse this trade keep optimizing toward spam.
Does this only work for medical practices?
The rubric changes; the system doesn't. A law firm scores case type and jurisdiction; a contractor scores project scope and service area; a SaaS company scores company size and use case. What's universal: individual scoring against explicit criteria, exclusion rules written down, and outcomes fed back into the ad platforms. We run variants of this across PPC engagements in every category we serve.
Can't smart bidding figure out lead quality on its own?
Not without being told what quality is. Smart bidding optimizes toward the conversions you define — if that's 'any form submission,' the algorithm dutifully hunts more form submissions, including the garbage. Offline conversion imports are how you teach it the difference. This was the core finding in the IVM engagement too: tracking defines the target; everything else optimizes toward whatever target you set. Set it carelessly and the machine executes your mistake at scale.