Sitting on a hospital subsidiary board as an operator

The first board meeting I attended as a director rather than a presenter, I did what I had always done. I read the deck the night before, marked the slides I disagreed with, and arrived ready to fix things. Twenty minutes in, the chair asked me a question I had not prepared for: not what I would do about the funnel, but whether the company had the right to be running that funnel at all, given whose patients the leads belonged to. I did not have an answer. I had spent a decade building demand engines. Nobody had ever asked me who owned the demand.

Hospital groups in India are spawning subsidiaries at pace: a digital health company, a diagnostics arm, a home care entity, a pharmacy platform, an insurance-facing services company. They are set up as separate legal entities for good reasons — capital, partners, regulation, a future listing or sale. The group needs directors on those boards who understand a digital business. The people who understand digital businesses inside the group are operators, and operators are the worst-prepared people for a board seat, because everything that made us good at our jobs is the thing a director is not supposed to do.

This is what I learnt about the seat, mostly by getting it wrong in the first two or three meetings.

The shift from executing to overseeing

A director does not run the company. You know this. You will still try to.

The trap is specific for a growth operator. You can read the subsidiary’s dashboard faster than anyone else in the room. You can see that the acquisition cost is climbing, that the app’s retention curve is flat, that the contact centre is closing enquiries at a rate the hospital side would never accept. Your instinct is to say so, in detail, and to prescribe. The moment you do, two things happen. The management team starts managing you instead of the business. And the rest of the board stops asking its own questions, because the digital expert has spoken.

The discipline I settled on: I am allowed to ask why, and I am allowed to ask what happens if. I am not allowed to say how. If the CEO of the subsidiary cannot answer why the acquisition cost is climbing, that is a board matter. If she can, and I disagree with her plan, I say so once, on the record, and then I let her run it. A director who cannot do that should go back to being an operator. There is no shame in it. There is considerable damage in pretending.

The other shift is time horizon. As an operator you are accountable for the quarter. As a director you are accountable for the entity — its solvency, its compliance, its reputation, its ability to exist in three years. Half the questions that matter at board level have no quarterly expression at all.

What you are actually there to protect

Strip away the ceremony and a director of a group subsidiary is protecting four things. The parent’s capital. The subsidiary’s independence, where the structure requires it. The patients whose data and money flow through it. And the group’s brand, which the subsidiary borrows and can spend faster than it can earn.

The last one is where a growth operator earns the seat. Most non-executive directors can read a balance sheet. Very few can read whether a digital health entity is quietly consuming the parent’s brand equity — running promotions the hospital would never run, making claims on a landing page that the clinical side would never sign off, or training a chatbot on content that has drifted from what the group actually promises. The board pack will not show you this. You have to go and look at the product yourself, as a patient would, before every meeting. I do. It takes an hour and it has surfaced more real issues than any slide.

Reading a digital business’s risks as a director

An operator reads a digital business through its funnel. A director has to read it through its failure modes. These are the ones I now look for first, in a rough order of how badly they hurt.

Demand that is not really its own. If most of the subsidiary’s revenue arrives because the parent’s hospitals route patients to it, the growth numbers are a transfer, not a market. Ask what the business looks like with the parent’s referrals removed. If nobody can answer, the entity has no idea what it is worth.
Unit economics that only work at a scale the market will not give. Home care, digital consults and diagnostics collection all look like software businesses in the model and like logistics businesses in reality. Ask for contribution margin per order in a Tier 2 city, not the blended national figure.
Regulatory exposure the founders treat as a detail. Telemedicine guidelines, drug scheduling for an online pharmacy, diagnostic lab accreditation, data protection obligations. In a hospital these live with a compliance function. In a two-year-old subsidiary they live with whoever remembered to read the circular.
Concentration in a single channel. One aggregator, one search platform, one insurer, one corporate client. Digital businesses drift into these because they work, and the board only notices when the terms change.
Technology debt dressed as velocity. A team shipping weekly is impressive until you ask who owns the data model and discover the answer is a vendor that could leave.

None of these will be on the agenda unless a director puts them there. Management, reasonably, brings the story it wants to tell. Your job is to make sure the story is not the only thing in the room.

Related-party questions between group companies

This is the part nobody briefs you on and the part where a growth operator is most exposed, because you are usually the person who set up the arrangement in your executive role that you are now supposed to scrutinise as a director.

The subsidiary uses the group’s brand. On what terms? Is there a licence, a fee, a set of usage rules, or just an assumption? The subsidiary receives leads from the group’s contact centre and website. Who pays for those leads, at what price, and who decided the price? The subsidiary’s patients get referred into the group’s hospitals. Is there a referral fee, and would it survive scrutiny from a regulator or a minority investor? The group’s IT team builds the subsidiary’s app. Is that a service agreement or a favour?

Every one of these is a related-party transaction. Every one of them needs to be documented, priced at something defensible, and disclosed. Most of them, in my experience, start as a conversation between two executives who trust each other and then sit undocumented for two years until an auditor, an investor or a tax officer asks.

A director’s question is not whether the arrangement is good for the group. It is whether it is fair to the subsidiary as an entity, and whether the subsidiary’s board would approve it if the parent were a stranger. When I sit on the subsidiary side, I have to argue against the terms my own function negotiated. The first time this happened I found it absurd. Now I think it is the whole point of having the seat.

The data question

Patient data is the related-party issue that will define these structures over the next few years, and most boards are not ready for it.

The group’s hospital holds clinical records. The subsidiary’s app holds behavioural data, payment data, sometimes consult records. The CRM sits somewhere in between. Everyone in the group wants a single view of the patient. The law increasingly wants to know which entity collected which data for which stated purpose, and whether the patient consented to it moving.

The board questions are plain. Which entity is the data fiduciary for each dataset? What consent did the patient give and to whom? Is data moving between the hospital and the subsidiary, and on what legal basis? Who can access it and is that logged? If the subsidiary is sold tomorrow, what data goes with it and what stays? If a regulator asks for the record of a data transfer, does one exist?

I have watched a management team present a beautiful personalisation roadmap built on combined hospital and app data, and I have had to be the person who asked whether the hospital had the right to share it. It is not a popular question. The room wants the roadmap. But a director who lets that go because the growth story is exciting has failed at the one thing the seat exists for.

The questions a good board member asks

Over time I have stopped preparing detailed critiques and started preparing questions. A short list, asked consistently, does more than a long list asked once. These are the ones that have earned their place.

What would this business look like without the parent? Revenue, brand, leads, people, systems.
Which number on this deck would you least like me to ask about, and why?
What did we promise a patient this quarter that we could not deliver?
Which single external party — platform, insurer, vendor, aggregator — could change the plan by changing its terms?
What has the compliance function said no to in the last quarter? If the answer is nothing, is anyone asking?
What are we spending on acquisition that we would stop if the parent stopped funding it?
Which decisions in this pack are actually decisions, and which are updates presented as decisions so that the board is on record?

The last one matters more than it looks. Management teams learn quickly that a decision taken by the board is a decision the board owns. A director who nods through a pricing change or a vendor contract disguised as an information item has taken on the risk without the deliberation.

What the seat costs you in your day job

Nobody tells you this either. Sitting on the subsidiary board changes your relationship with its management team, and they are often people you work with daily in your executive role. The CEO of the subsidiary may report to a group executive who is your peer. Now you are one of her governors. Every operational conversation you have with her carries a second meaning.

The way I handle it is to be explicit. In the board room I am a director and I will ask hard questions. Outside it, I am a colleague and I will help. I try never to bring an operational disagreement into the board room dressed as a governance concern, and I try never to use the board seat to get an operational outcome I could not get as a peer. Both temptations are real. Both erode the seat.

The other cost is time. A board seat done properly takes a day of preparation per meeting, plus committee work, plus the hour you spend as a mystery patient. Done badly it takes ninety minutes and a signature. The group will not notice which one you are doing until something goes wrong.

If you are taking a seat next quarter

The order of operations that would have saved me the first two meetings.

Read the articles of association, the shareholders’ agreement if there is one, and the last audit report before you read any deck. Know what the entity is and who can do what to it.
List every arrangement between the subsidiary and the group — brand, leads, referrals, shared services, data, people on secondment. Ask for the document behind each one. Where there is no document, that is your first agenda item.
Use the product as a patient. Book something, cancel something, complain about something. Do it again before every meeting.
Ask management for the three numbers they run the business on, and agree with the board that those three will be on the first page of every pack.
Decide, in writing to yourself, where your executive role and your director role conflict. Declare the conflicts to the chair before someone else finds them.
In your first meeting, ask more than you say. In your second, ask the questions from the list above. By the third, the management team will have started preparing for them, which is the outcome you want.

Governance is not the opposite of growth. It is what growth looks like when someone has to be accountable for it after the operator has moved on.

Questions people ask

What does a director on a hospital subsidiary board do differently from an operator?

A director does not run the company, and an operator will still try to. The trap for a growth operator is that you can read the subsidiary’s dashboard faster than anyone in the room, and the instinct is to prescribe. The moment you do, management starts managing you instead of the business and the rest of the board stops asking questions. The discipline I settled on: I may ask why and what happens if. I may not say how. Disagree once, on the record, then let the CEO run it.

What is a director of a hospital group subsidiary actually there to protect?

Four things. The parent’s capital. The subsidiary’s independence, where the structure requires it. The patients whose data and money flow through it. And the group’s brand, which the subsidiary borrows and can spend faster than it can earn. The last is where a growth operator earns the seat, because very few non-executive directors can tell whether a digital health entity is quietly consuming the parent’s brand equity through promotions, landing-page claims or a chatbot the clinical side would never sign off.

What are the biggest risks in a hospital group’s digital subsidiary?

Demand that is not really its own — revenue that arrives because the parent routes patients to it. Unit economics that only work at a scale the market will not give, because home care and diagnostics look like software in the model and logistics in reality. Regulatory exposure the founders treat as a detail. Concentration in one aggregator, insurer or corporate. And technology debt dressed as velocity, where the data model belongs to a vendor who could leave. None of these reach the agenda unless a director puts them there.

What counts as a related-party transaction between a hospital and its subsidiary?

Almost every arrangement that started as a conversation between two executives who trusted each other. The subsidiary uses the group’s brand — on what terms? It receives leads from the group’s contact centre — who pays, at what price? Its patients are referred into the group’s hospitals — is there a fee, and would it survive a regulator or minority investor? The group’s IT team builds its app — service agreement or favour? Each needs to be documented, priced defensibly and disclosed, usually before an auditor asks.

Who owns patient data when a hospital shares it with a subsidiary app?

That is the question most boards are not ready for. The hospital holds clinical records; the app holds behavioural, payment and sometimes consult data; the CRM sits between. Everyone wants a single view of the patient. The law wants to know which entity collected which data for which purpose, and whether the patient consented to it moving. A director must ask which entity is the data fiduciary for each dataset, on what legal basis data moves, whether access is logged, and what goes with the subsidiary if sold.

What questions should a board member ask a digital health subsidiary?

A short list asked consistently does more than a long list asked once. What would this business look like without the parent? Which number on this deck would you least like me to ask about? What did we promise a patient this quarter that we could not deliver? Which single external party could change the plan by changing its terms? What has compliance said no to lately? What acquisition spend would stop if the parent stopped funding it? And which items are actually decisions rather than updates dressed as decisions?

How much time does a hospital subsidiary board seat actually take?

Done properly, a day of preparation per meeting, plus committee work, plus an hour spent using the product as a patient before every meeting. Done badly, ninety minutes and a signature. The group will not notice which one you are doing until something goes wrong. The preparation is not reading the deck and marking slides to fix — I did that in my first meeting and it was the wrong preparation. It is reading the constitutional documents, listing the intercompany arrangements and preparing questions rather than critiques.

How does a subsidiary board seat change your relationship with its management team?

Every operational conversation acquires a second meaning, because the subsidiary CEO you work with daily is now someone you govern. The way I handle it is to be explicit: in the boardroom I am a director and will ask hard questions; outside it I am a colleague and will help. I try never to bring an operational disagreement into the boardroom dressed as governance, and never to use the seat for an operational outcome I could not get as a peer. Both temptations are real and both erode the seat.

What should you read before your first meeting on a hospital subsidiary board?

The articles of association, the shareholders’ agreement if there is one, and the last audit report — before any deck. Know what the entity is and who can do what to it. Then list every arrangement between the subsidiary and the group: brand, leads, referrals, shared services, data, people on secondment. Ask for the document behind each one. Where there is no document, that is your first agenda item. And decide, in writing to yourself, where your executive role and director role conflict, then declare it to the chair.

Why should a director use the subsidiary’s product as a patient before every board meeting?

Because the board pack will not show you whether the entity is spending the parent’s brand. Book something, cancel something, complain about something. In an hour you see the promotions the hospital would never run, the claim on a landing page the clinical side never approved, the chatbot trained on content that has drifted from what the group promises. It has surfaced more real issues for me than any slide, and management cannot prepare a deck around a director who has just tried to get a refund.

How can an investor tell a subsidiary’s real revenue from transfers by the parent hospital?

Ask what the business looks like with the parent’s referrals removed — revenue, brand, leads, people, systems. If most revenue arrives because the group’s hospitals route patients in, the growth numbers are a transfer, not a market, and the entity has no idea what it is worth. Then ask for contribution margin per order in a Tier 2 city rather than the blended national figure, and for the share of acquisition spend that would stop if the parent stopped funding it. Those three answers usually settle the valuation conversation.

When should a growth operator decline a seat on a hospital subsidiary board?

When you cannot stop saying how. A director who cannot state a disagreement once and then let management run it should stay an operator — there is no shame in it and considerable damage in pretending. Decline also when the conflict is unmanageable: if you negotiated the intercompany terms in your executive role and cannot honestly argue against them on the subsidiary side, the seat is compromised. And decline if you cannot give it a day of preparation per meeting. A signature is not governance.

What did you get wrong in your first meetings as a subsidiary director?

I arrived ready to fix the funnel, and the chair asked whether the company had the right to be running that funnel at all, given whose patients the leads belonged to. I had spent a decade building demand engines and nobody had ever asked me who owned the demand. I also found it absurd, the first time, to argue against terms my own function had negotiated with the subsidiary. Now I think that is the whole point of having the seat.

What is a governance risk in a subsidiary board pack that looks like an information item?

A pricing change or a vendor contract presented as an update rather than a decision. Management teams learn quickly that a decision taken by the board is one the board owns, so items arrive as information so the board is on record without having deliberated. A director who nods one through has taken on the risk without the discussion. Ask, every meeting, which items are actually decisions. By the third meeting the management team will have started preparing for that question, which is the outcome you want.