What is Lead Source Quality (LSQ)?

LSQ is a 0-100 score per Meta ad set blending consult-show rate, treatment-conversion rate, average ticket size, and refund/cancellation rate from the clinic CRM. It tells you which ad sets deliver revenue-grade leads, not just cheap form-fills.

Why is CPL misleading on its own?

A ₹420 IVF lead that no-shows is more expensive than a ₹1,400 lead that consults and converts. CPL alone optimises for friction-free junk; LSQ optimises for revenue. We routinely see a 2-3x gap between best CPL ad set and best LSQ ad set in the same account.

How long does LSQ need to stabilise?

For IVF and hair transplant: 21-28 days because consult-to-treatment cycles are long. For dental aligners and derm: 10-14 days. We only act on LSQ scores once an ad set has 25+ leads with full CRM-stage outcomes recorded.

Meta Ads · LSQ Scoring · 2026

LSQ + Meta Integration — Scoring Lead Source Quality So Spend Follows Revenue, Not Form-Fills

Published 27 June 2026 · ICG Editorial · 6 min read

Most healthcare clinics optimise Meta ad sets on CPL. Meta Catalyst IQ optimises on Lead Source Quality — a stitched score that pulls CRM consult-show, treatment-conversion, and ticket size back to the ad-set level. The gap between best CPL and best LSQ ad set in the same account is often 2-3x. Spending on the wrong winner for 90 days is how clinics burn 18-22% of their annual ad budget on junk.

Why CPL alone breaks healthcare ad ops

A ₹420 IVF lead that doesn't pick up the phone is not a cheap lead. It's an expensive non-lead. But Meta's optimiser sees a "conversion" the moment the form is submitted, and so does your weekly report. Three months in, the ad set with the lowest CPL is also the ad set producing the lowest revenue — and nobody noticed because the dashboard ended at form submission.

The Indian healthcare context makes this worse. Lead-form completion rates are inflated by autofill. WhatsApp-first behaviour means many leads never call back. Tier-2 city users frequently fill multiple clinic forms in one sitting. The friction layer between "form" and "consult" is wider here than almost anywhere else.

The LSQ scoring formula Meta Catalyst IQ runs

Lead Source Quality is a weighted 0-100 score per ad set, calculated nightly from CRM webhook data:

Component	Weight	What it captures
Consult-show rate	30%	Lead actually attended consult (in-clinic or video)
Treatment-conversion rate	30%	Consult converted into paid treatment
Average ticket size	20%	Revenue per converted lead, normalised against specialty median
Refund / cancellation rate	10%	Negative weight: how many bookings collapsed
Time-to-decision	10%	Faster decisions = higher intent = higher LSQ

Each component is normalised against the specialty's 90-day rolling median, so a dental aligner ad set isn't unfairly compared to an IVF ad set with a 6-week sales cycle.

A worked example from an Indian fertility chain

An anonymised fertility chain running Meta across Delhi NCR, Bengaluru, and Hyderabad in Q1 2026 had four "winning" ad sets by CPL. After 35 days of LSQ stitching, the picture inverted.

Ad set	CPL	Consult-show	Tx-conv	LSQ score	True CAC
NCR · Cost ladder · Broad	₹412	22%	9%	34	₹20,800
NCR · Doctor video · 30-42	₹980	61%	27%	82	₹5,950
BLR · Testimonial · LAL 1%	₹740	54%	22%	74	₹6,220
HYD · Discount · Broad	₹385	18%	6%	26	₹35,650

The CPL leaderboard told the team to pour spend into the Hyderabad discount campaign. LSQ told the truth: that campaign was producing leads at ₹35,650 per actual treatment — 6x worse than the doctor-video ad set the team had nearly killed.

How the integration plumbing works

Meta lead → CRM: Catalyst IQ pulls Meta Lead Ads + on-Meta WhatsApp leads via the Marketing API, hashes PII, and posts into the clinic CRM with ad-set ID retained as a custom field.
CRM stage events → Catalyst IQ: Each stage transition (Contacted → Consult Booked → Consult Showed → Treatment Paid → Refunded) fires a webhook back, tagged to the originating ad-set ID.
Nightly LSQ recompute: A scheduled job recalculates LSQ for every ad set with 25+ leads and outcomes settled to T-7.
Action Centre surfacing: Ad sets that drop below LSQ 40 trigger a "Scale-or-kill" diagnostic; ad sets above LSQ 70 trigger a "Scale candidate" recommendation.

The mistakes most agencies make

Acting too early

LSQ needs 21-28 days of CRM-stage settlement for IVF and hair transplant, 10-14 for dental aligners and derm. Acting on a 7-day LSQ score is no better than acting on CPL — the consults haven't happened yet.

Not normalising by specialty

A 22% consult-show rate is terrible for dental aligners (where median is 55%+) but excellent for IVF (where median is 28-32%). Without specialty normalisation, you'll euthanise good IVF ad sets and protect bad dental ad sets.

Treating refunds as silent

Refund and cancellation events are the truest negative signal. A 4-week treatment paid then refunded at week 6 means the lead was wrong-fit, the ad creative was misleading, or the consult oversold. All three need ad-account corrections.

The insight

CPL is a vanity metric for healthcare ad sets that lack CRM stitching. The clinic that wins isn't the one with the cheapest leads — it's the one whose ad sets are ranked by Lead Source Quality and whose budget reallocations happen weekly based on consult-show and treatment-conversion data flowing back from the CRM.

Want LSQ scoring on your Meta ad accounts?

We'll plumb CRM stage events back to your Meta ad sets, rank them by Lead Source Quality, and run reallocations weekly. Audit is free.

Book a free audit → WhatsApp ICG