Centralising marketing across hospital units

Most multi-unit hospital groups in India run marketing the way they acquired hospitals: one unit at a time. Each hospital has its own marketing manager, its own agency, its own idea of what the brand looks like, and its own relationship with the unit head who signs the cheques. The group logo is the only thing they share.

Centralising that function is one of the highest-return decisions a group can make. It is also one of the most resisted, because every unit head experiences it as a loss of control. This is how to think about it, what it actually changes, and where centralisation should stop.

What decentralised marketing costs you

The costs are real but scattered, which is why they rarely appear on one slide.

Duplicate spend on the same keywords. Two hospitals in the same city bidding against each other for “best cardiologist near me” is not a hypothetical. It happens in every group where units run their own paid search. You pay a premium to compete with yourself.

Brand fragmentation. Different fonts, different taglines, different doctor-photo styles, different tones on social media. To a patient comparing hospitals on a phone screen, this reads as several small brands rather than one large one. You lose the trust premium that the group name should carry.

No shared learning. The oncology campaign that worked in one city never reaches the unit three hundred kilometres away because there is no mechanism for it to travel. Every unit relearns the same lessons at full price.

Weak negotiating position. Six agencies each billing a unit is six small contracts with no leverage. One agency relationship for the group is a different conversation.

Unmeasurable attribution. Each unit’s agency reports its own numbers, in its own format, to its own unit head. Nobody at the group can answer the question “what is our cost per new patient, and which channel is moving it?”

What centralisation actually means

It does not mean a head-office team that decides everything. That model fails within a year because it cannot know what is happening on the ground in each city. Centralisation that works separates three layers.

Layer 1: Platform and standards (fully central)

• Brand identity, templates, tone of voice, doctor photography standards
• Website architecture, CRM, marketing automation, analytics stack
• Paid media buying and bid management across all units
• Agency contracts and rate cards
• Attribution model and the reporting format every unit uses

These are things where one good decision beats twenty local ones, and where consistency itself has value.

Layer 2: Programmes (central design, local execution)

• Service-line campaigns: oncology, cardiac, orthopaedics, maternity
• Seasonal and awareness calendars
• Doctor onboarding and profile building
• Referral and corporate outreach playbooks

The group designs the campaign once. Each unit adapts the doctor names, the language, the local proof points, and runs it.

Layer 3: Local presence (fully local)

• Community events, camps, local partnerships
• Google Business Profile management and review responses
• Relationships with local referring doctors and clinics
• Regional-language content and local media

Head office should set the standard for these and then get out of the way.

The mistake groups make is centralising Layer 3 or leaving Layer 1 local. Both fail.

The Indian specifics

Language is a structural issue, not a translation task. A group spanning Hyderabad, Chennai, Kolkata and Delhi is marketing in at least five languages. Central teams default to English and Hindi and lose the regional patient. The central layer should own the templates and the approval workflow; the local layer must own the language.

Unit heads are often clinicians or hospital administrators with P&L accountability. They are not marketers, and they know it, but they carry the number. Centralisation only works if the group marketing function commits to unit-level targets and reports against them monthly. If the unit head cannot see their own funnel, they will rebuild their own team within two quarters.

Referral doctors are a marketing channel in India in a way they are not elsewhere. A significant share of planned admissions arrive via a referring GP or a smaller nursing home. This channel is inherently local and relationship-driven. Centralise the CRM that tracks it. Do not centralise the people who work it.

Doctor brands compete with the hospital brand. A star surgeon with a large personal following will resist a central brand template that makes them look like every other doctor. The answer is not to fight; it is to give star doctors a premium tier within the brand system, so that promoting them also promotes the group.

Sequencing the transition

The order matters more than the org chart.

1. Attribution first. Before you move a single person, get every unit reporting the same funnel in the same format: enquiries, appointments, first visits, admissions, by channel. This takes one quarter and it creates the shared language for everything after. It also shows unit heads what they are currently getting for their money, which is usually the moment resistance softens.
2. Paid media second. Consolidate the ad accounts under one structure with unit-level campaigns. This is where the fastest, most visible saving lives, and it does not require anyone to change reporting lines.
3. Agencies third. Once you control the accounts, consolidating agencies is a commercial exercise rather than a political one.
4. Brand and templates fourth. Roll out the identity system with the tools already in place, so units adopt it because it is easier, not because they were told to.
5. People last. Only now decide who reports where. By this stage the unit marketing managers are already working inside central systems and the reporting-line change is a formality.

Groups that start with the org chart spend a year fighting and never reach step one.

The governance that keeps it working

A monthly unit review with three fixed items: funnel performance against target, spend against plan, and one thing the unit needs from the centre. Fifteen minutes per unit. The point is not the meeting; it is that unit heads know they have a forum and that the centre is accountable to them, not the other way round.

A quarterly programme review where each service line’s campaign is examined across all units. This is where shared learning actually happens, because one unit’s outlier result becomes everyone’s next quarter.

A single dashboard that unit heads and the group CEO both look at. If they are looking at different numbers, centralisation has not happened.

What you give up

Speed on hyper-local decisions, at least initially. A unit that used to change its billboard creative in a day now goes through a template process. Accept this and shorten the process rather than reopening the principle.

Some unit-level talent. Marketing managers who joined for autonomy may leave. The ones who stay tend to be the ones who wanted better tools.

A period of worse-looking numbers. Consistent attribution usually reveals that the previous, self-reported unit numbers were inflated. Warn the CEO before this happens.

What you get

A defensible cost per patient across the group. A brand that reads as one institution. The ability to launch a service line in every city in the same month. Agency and media rates that reflect your true scale. And a marketing function that can be handed a growth target and be held to it.

The budget model that survives the transition

Centralisation collapses if the money moves before the accountability does. The model that holds up in Indian hospital groups is a split budget with a transparent chargeback.

The group holds a central budget for Layer 1: platforms, brand, agency retainers and group-level paid media. This is funded as a group overhead, not allocated to units, because units cannot influence it and should not be asked to defend it.

Each unit holds a local budget for Layer 3, sized by its revenue and growth target, and spends it with local discretion inside the brand standards. This preserves the unit head’s sense of ownership and keeps local relationships funded.

Layer 2 programme spend is where the chargeback lives. The centre designs and runs the campaign; units are charged their share of media and production in proportion to the leads and appointments attributed to them. Because the attribution model is shared, the charge is defensible and the unit head can see exactly what they paid for. This is the mechanism that turns central marketing from a cost centre that units resent into a service that units request.

The agency question

Consolidating from many agencies to one or two is a saving on paper and a risk in practice. A single national agency will service your metro units well and your tier-two units poorly, because that is where its people are. The pattern that works is one lead agency for strategy, brand, digital platforms and paid media, plus a small panel of regional agencies for local-language creative and on-ground activation, contracted centrally on a common rate card but tasked locally.

Bring performance media in-house within the first two years if you can. It is the largest controllable spend, the skills are available in the Indian market, and the data belongs inside the group rather than inside an agency’s account structure. The lead agency then becomes a creative and strategic partner rather than a media intermediary, which is a healthier relationship for both sides.

A note on timing

The best moment to centralise is when the group is about to add units, not after. A new hospital launched into an already-central marketing system starts with the right templates, the right attribution and the right agency from day one. A new hospital launched into a decentralised system creates one more fiefdom to unwind later. If a launch is on the roadmap, sequence the centralisation to land before it.

If you’re planning this for next year

• Start with the attribution model. Nothing else works without it.
• Decide the three layers explicitly and write them down. Ambiguity is where the fights start.
• Give unit heads a monthly forum and a dashboard before you take anything away from them.
• Consolidate paid media early. It funds the rest.
• Move people last.

Centralisation is a systems change dressed up as a reporting-line change. Treat it that way and the reporting lines take care of themselves.