Healthcare Meta Ads: Broad vs Narrow Targeting in India (2026 Playbook)

The "broad targeting won" argument has become lazy shorthand for "we stopped thinking about audiences." In healthcare specifically, that shortcut costs money. Broad works in some places, narrow still wins in others, and the difference is not ideology — it is arithmetic. This playbook walks through the audience-sizing decision the way we make it inside Meta Catalyst IQ, across 23+ healthcare accounts and ₹9.14Cr of monthly optimised spend. When to open the audience, when to keep it tight, and how to read Meta's "Audience Size vs Efficiency" verdict without getting fooled by the ranking bar.

What broad and narrow actually mean in 2026

Terminology has drifted. In 2026, "broad" usually means location + age + gender only, plus a couple of language layers, with Advantage+ audience turned on. "Narrow" means location + age + gender + interest stack + optional lookalike layer, with Advantage+ audience turned off or restricted. On Meta's current interface, "Advantage+ audience" is not a single-click toggle everywhere — depending on account type and campaign objective, it appears differently. The important part is not the button. It is whether Meta is allowed to expand beyond your defined audience or not.

When broad targeting wins

Broad wins when three conditions overlap. First, the addressable audience in your city is large — over 5 lakh in the age-gender-language block. Second, your creative has clear self-qualifying language in the first 3 seconds, so wrong audiences self-eliminate. Third, your consult team can handle a mixed lead pipeline without breaking. Under these conditions, broad usually pulls a 15–30% cheaper CPL than a narrow audience of the same spend level, because Meta's optimisation has more surface to search. IVF in Delhi NCR, dermatology in Bengaluru, and aesthetic in Mumbai all typically hit these conditions. Broad works there.

When narrow still wins

Narrow beats broad when the addressable audience is small, when creative self-qualification is weak, or when the consult team cannot triage. In tier-2 and tier-3 cities where the total in-market audience is under 2 lakh, narrow interest stacking still outperforms because broad simply repeats to the same 40,000 people and burns frequency fast. For specialties where the patient decision is expensive and consultative — bariatric, oncology, IVF at premium price points, joint replacement — narrow with intent-signal interest layers converts materially better. If you are running a ₹3.5 lakh cardiac spend in Coimbatore, do not open to broad; you will get 30 wrong-fit leads a day and your consult team will burn out.

The audience size vs efficiency verdict

Every account in Meta Catalyst IQ carries an Audience Size vs Efficiency verdict, computed monthly. It looks at total addressable audience across all live ad sets, the frequency curve, the CPL trajectory, and the consult-conversion mix. The verdict comes out as one of four labels: Too Tight (frequency > 4, audience below 200,000), Balanced (frequency 1.5–3, audience 300,000–1.5M), Over-Opened (frequency < 1.5 but CPL climbing, consult conversion below account median), and Fragmented (ten or more ad sets serving fewer than 20 conversions each). Each verdict has a distinct fix. Too Tight means add lookalikes or open geography. Over-Opened means add an interest layer or restrict expansion. Fragmented means consolidate ad sets — the classic post-iOS trap.

The Fragmented Account trap

The most common broken pattern we inherit from new clients is Fragmented. Someone spent a year building fifty ad sets for granular targeting — every micro-interest, every city sub-area, every age bucket. Post iOS14 signal loss, none of them get enough conversions per week to exit learning phase. CPL is high, learning phase is never leaving, and the answer is not more creative — the answer is consolidation. Merge to five to seven ad sets, each with 20+ weekly conversions, and CPL usually drops 25–40% in three weeks. This is one of the largest recurring wins we log in the Money Wastage Cleanup module.

Lookalikes still work — but differently

Lookalikes have not died. They have quietly changed roles. In 2020 a 1% lookalike of your customer list was the primary audience. In 2026 lookalikes are a seeding layer inside broad or a narrow-in-narrow layer for tier-2 cities. What still works: 1–3% lookalikes of consulted patients (not top-of-funnel leads), 1% lookalikes of high-LTV segments, and lookalikes of website purchases on ecommerce-adjacent healthcare (aligners, cosmetic packages). What has stopped working: lookalikes of Meta lead form fills where the fill-to-consult conversion is under 20% — the seed is too noisy and Meta will find you more noise.

Interest stacks that still outperform in India

Broad Advantage+ audiences work well in metros; interest stacks still outperform in tier-2 and tier-3. Real interest layers we still use with measurable lift:

• Parenting communities for IVF in tier-2 cities
• Skincare and beauty routine communities for derm below age 30
• Bridal and event-adjacent interests for aesthetic in wedding season
• Cardiac patient support communities for hospital lead-gen in tier-2
• Higher-income proxy interests (luxury retail, business travel) for premium price positioning

Do not over-stack. Two to three interest layers is the ceiling. Beyond that, ad set audience shrinks below Meta's optimisation floor and CPL climbs.

Geo-targeting inside a city — the pin drop question

Should you pin drop to a 5-kilometre radius around the clinic, or run a whole-city audience? Answer: it depends on the specialty and the patient travel radius. Dental primary and aesthetic patients travel 3–8 kilometres. IVF and oncology patients travel 15–40 kilometres and often across cities. Cardiac emergency will travel to the nearest ICU. For dental and aesthetic, run pin-drops around the clinic and each satellite. For IVF and oncology, run whole-city and let the audience self-select on booking. Wrong geo tightness is a common source of Too Tight verdicts in the portfolio.

Age and gender — two hard rules

Two rules that rarely change. First, gender-splitting IVF audiences into female-only is efficient early but leaves 30% of the actual decision layer out — men are often the primary financial decision-maker for IVF spend in Indian households, and creative that speaks to couples converts better than creative that speaks to women only. Run mixed-gender for IVF from month two onwards. Second, do not tighten age windows below 15 years unless the specialty demands it (paediatric, geriatric). Meta's optimisation punishes tight age windows more than most managers realise.

Reading the frequency curve properly

Frequency below 1.5 means the audience is too big or the spend is too low. Frequency between 1.5 and 3 is the healthy band. Frequency above 3.5 in a metro or 2.8 in a tier-2 means audience saturation is starting to damage CPL. The mistake teams make: they read frequency at the campaign level. Read it at the ad set level, weekly, and act on the outlier ad sets. The 2-Day Comparative inside Meta Catalyst IQ flags any ad set whose frequency climbs 20% week over week — that is the early fatigue signal, worth acting on before CPL confirms.

Notes on Advantage+ audience expansion

Advantage+ audience is not the same as Advantage+ Shopping Campaigns. For healthcare lead-gen, Advantage+ audience expansion is a tool, not a strategy. Turn it on when your baseline audience is small and your creative self-qualifies. Turn it off when you are running in a tier-2 city, when creative is generic, or when consult-conversion drops after enabling it. Measure the before-and-after over two full weeks — one week is not enough because auction dynamics shift on the seven-day cycle.

The exclusions that matter — and the ones that hurt

Exclusion lists shape the audience as much as inclusions do. Two exclusions almost every healthcare account should run: existing patients (via customer list) to stop wasting spend re-acquiring people you already treat, and current employees / staff to prevent the internal engagement that skews CTR. Two exclusions we usually turn off: excluding all past form-fillers (kills valuable retargeting), and excluding based on very broad interests (usually cuts more real prospects than noise). Every three months, audit the exclusion stack — inherited accounts often carry exclusions that made sense two campaigns ago and quietly bleed audience today.

Language layers and placement selection

Two often-ignored targeting levers. First, language: adding regional language layers (Tamil, Telugu, Marathi, Malayalam, Kannada, Bangla) alongside English materially opens audience in tier-2 cities without sacrificing quality — patients who read regional-language content still consult in the same clinics. Second, placement: Advantage+ placements is the default, but for consultative specialties (cardiac, IVF, oncology) manual placement selection favouring Facebook Feed and Instagram Feed over Reels and Stories often improves consult-conversion, because the reader is in a longer-attention mode. Test both — the default is not always the answer.

How the audience decision changes with brand maturity

A new brand in a new city needs narrow interest stacking to build early signal — Meta has nothing to learn from, so give it a shape. A brand with 12 months of clean CAPI events and a strong lookalike seed can open to broad and let Meta's optimisation carry the load. A brand with 3 years of history in a specialty typically runs a mix — broad for prospecting, lookalikes for scaling, narrow retargeting for BOFU. The right targeting mix is not universal; it is a function of how much your brand has already taught Meta about who your patients are. This is why audience decisions should be revisited quarterly, not set once and forgotten.

Powered by Meta Catalyst IQ — the decision engine behind every Meta ad ICG runs

ICG built Meta Catalyst IQ because most Indian healthcare brands running Meta ads waste 30–50% of budget without knowing it. It's the diagnosis + decision layer above Ads Manager — Hygiene Factors 12-point checklist, Naming Intelligence (surfaces conflicts costing ₹50K–₹2L/account/month), Creative Scoring Matrix (Core Performer / Scalable / Getting Started / Review), 2-Day Comparative, SLC Framework, Money Wastage column in ₹.

• Master Dashboard — 23+ accounts, ₹9.1Cr+ spend/mo optimised, ₹1,581 blended CPL vs ~₹3,200 market benchmark.
• Diagnose → Optimise → Grow — daily hygiene checks, weekly creative scoring, monthly money wastage cleanup.
• CPQL Engine — cost per qualified lead (not just cost per lead) at ad-set level. Try the interactive CPQL calculator.
• Portfolio benchmarks — IVF ₹632, derm ₹520–1,180, dental ₹620–1,800, aesthetic ₹400–900, hospital cardiac ₹3,200.

Included free with every ICG Meta ads or Performance Marketing engagement (Starter ₹20,000/-/month tier and above). Not sold standalone. Book a free 48-hour Meta ad diagnostic or WhatsApp us.

Frequently asked questions

Is broad targeting always better in 2026?

No. Broad wins in metros with large addressable audiences and creative that self-qualifies. In tier-2/3 cities with small addressable audiences or consultative specialties, narrow interest stacks still outperform.

How many ad sets should a healthcare account run at once?

Five to seven is a healthy default. Fewer than three under-tests audience hypotheses; more than ten usually fragments and prevents ad sets from exiting learning phase.

Do lookalike audiences still work?

Yes, but with cleaner seeds. Lookalikes of consulted patients or high-LTV segments outperform lookalikes of raw form fills. Seed quality is more important than percentage.

What frequency should we act on?

Above 3.5 in a metro or 2.8 in a tier-2 city, refresh creative or open the audience. Below 1.5, expand or increase spend — Meta is not seeing enough auctions to optimise.

Should we pin drop to a small radius around the clinic?

For dental and aesthetic, yes — patients travel 3–8 km. For IVF, oncology, and cardiac, run whole-city and let booking self-select — those patients travel much further.

What is the biggest targeting mistake ICG sees in inherited accounts?

Fragmentation. Fifty ad sets, each with too few conversions to exit learning phase. Consolidation to 5–7 ad sets typically drops CPL 25–40% inside three weeks.

Related reading

• Lookalike audiences for healthcare Meta ads — 2026 playbook
• The 5-zone creative testing framework for healthcare Meta ads
• Meta CAPI and server-side events for healthcare marketers
• EMQ score optimisation for signal quality
• Hook frameworks for the first 3 seconds

Authority reading: Meta's documentation on audience targeting best practices.