Linkedin A1 Beacon Capi Healthcare Standard

Length: ~1,300 words

When I joined a call with a hospital group CMO last year, she opened with a direct statement: "Our digital agency has CAPI deployed. Our Meta EMQ is 5.2. We are told this is good."

I explained why it was not good enough for healthcare — and why the standard we hold ICG's engagements to is an EMQ of 7.5 or above, consistently, within 6 weeks of deployment.

Here is the technical case, and why it matters commercially.

What Meta Event Match Quality actually measures

Meta's EMQ score (1–10) measures how completely a conversion event sent to Meta can be matched to a Meta user. A higher EMQ means Meta can more precisely identify who converted — and therefore find more people who resemble them.

An EMQ below 4.5 means Meta's algorithm is operating in a data environment where it cannot reliably identify the converting audience. It is, in effect, guessing. Above 7.0, the algorithm has sufficient matching signal to identify genuine patterns. Above 8.0, the matching is high-quality.

For e-commerce brands with millions of conversion events, even a moderate EMQ produces adequate algorithm performance — the volume of data compensates for the quality gap.

For healthcare, this is not true. A specialty clinic generating 150–300 monthly consultation enquiries from Meta does not have the volume buffer that compensates for low-quality matching. Every conversion event needs to match correctly. EMQ 5.2 is inadequate.

Why healthcare generates lower EMQ by default

Healthcare Meta accounts face specific EMQ challenges that consumer brands do not:

Patient privacy behaviour: Healthcare enquirers use private browsing and iOS devices at above-average rates, driven by the sensitivity of the health information they are sharing. This suppresses client-side pixel data capture — the standard source of the customer data that feeds EMQ.

Long consideration cycles: A patient who clicks a Meta ad for IVF in January and enquires in April (90-180 day consideration cycle) produces a conversion event 90-180 days after the original click. Standard pixel-based tracking, with a 7-day conversion window, captures none of these. EMQ is built from zero matched events.

WhatsApp as primary conversion channel: In most Indian healthcare markets, the conversion happens on WhatsApp — a click that exits the browser environment. Without server-side capture of the WhatsApp click event, the conversion is invisible to the pixel, invisible to Meta's algorithm, and invisible to EMQ calculation.

What Beacon does differently

ICG's Beacon CAPI addresses all three EMQ challenges specifically for healthcare:

Server-side capture of WhatsApp click events — before the browser is exited. The patient's WhatsApp click is recorded by Beacon at the server level, along with any available user identifiers (hashed phone, hashed email from the form they submitted). This event is transmitted to Meta CAPI with full attribution — bypassing the iOS and browser limitations.

180-day attribution window configuration for IVF and other long-consideration specialties. Beacon logs the original click event and attributes it to a consultation booking 90-180 days later — sending the matched offline conversion event to Meta through the Conversions API.

DPDP Act 2023-compliant data hashing — all user identifiers are SHA-256 hashed before transmission to Meta. The matching is done by Meta on their side using hashed comparison. Raw personal data is never transmitted.

The result: Meta EMQ across ICG's Beacon-deployed healthcare accounts moves from the typical 3.8–4.4 (pixel-only baseline) to 7.5–8.5 within 6 weeks. CPM falls 25–38% as the algorithm finds higher-quality audiences. CPQL falls 30–45%.

Why "CAPI deployed" is not the same as "Beacon deployed"

The hospital group CMO whose agency had "CAPI deployed" had a partial implementation: form submission events sent server-side, but WhatsApp click events still captured client-side (pixel only), and no offline conversion import for consultation attendance. The EMQ of 5.2 reflected this partial implementation — better than pure pixel, not good enough for reliable algorithm optimisation.

The specific standard ICG holds:

1. WhatsApp click events captured server-side (not pixel)
2. Consultation booking events sent as offline conversions via CAPI (not estimated from form submissions)
3. Consultation attendance events sent as offline conversions (this is the signal that teaches the algorithm what a converting patient looks like — not just who clicks a form)
4. 180-day attribution window for IVF and elective surgery specialties
5. DPDP-compliant SHA-256 hashing on all transmitted identifiers
6. EMQ ≥ 7.5 verified in Meta Events Manager within 6 weeks of deployment

When all 6 elements are in place, Meta's algorithm has what it needs to find the right patients. The CPQL falls. The attribution improves. The marketing P&L becomes readable.

EMQ 5.2 is not good enough. 7.5 is the floor. This is not a perfectionist standard — it is the minimum required for reliable healthcare algorithm performance.

→ Revenue Operations for Healthcare