Building a digital team inside a hospital group

The first thing you will be asked for is an organisation chart. Resist it for a month.

When digital landed on my desk at a large hospital group, the expectation was structure — boxes, titles, reporting lines, a slide the board could absorb in twenty seconds. What the group actually needed was three people who could do the work while the rest of us worked out what the work was. An org chart drawn in week one encodes your assumptions about a business you do not yet understand, and in a hospital group those assumptions are wrong in expensive ways.

A hospital group is not one business. It is a collection of units with their own P&L, their own referral geography, their own specialty mix and their own opinion on whether you are help or overhead. Whatever team you build has to function inside that reality, not above it. That single fact shapes your hiring order more than any functional logic about search, performance marketing or product.

What the job actually is in the first year

It is plumbing. Not growth, not brand, not transformation. Plumbing.

In the first ninety days at the group, we counted the phone numbers the organisation was publishing to patients across its units, its campaigns, its signage and its listings. The number was absurd. Several of them rang in departments that no longer existed in that form. Several rang in a unit that had moved floors. One rang a mobile phone belonging to a front-office executive who had left.

Then we counted map listings per unit. Most units had more than one, created at different times by different people, with different timings, and in two cases with the wrong entrance marked — which matters at five in the morning when a patient is looking for an emergency gate. Doctor profiles existed in three places with three different sets of qualifications. Nobody could say with confidence who held the administrative login to the group website.

That is the real first-year job, and it dictates the first hire. You are not hiring someone to build a demand engine. You are hiring someone to make the basic surfaces of the organisation true, and to keep them true while forty-odd stakeholders keep changing them.

The first hire: digital operations, not marketing

Hire someone whose instinct is to open a tracker, not a deck.

The profile that works is a delivery or operations person — mid-sized e-commerce operations, an agency delivery lead, someone who has run a migration with a stubborn vendor. They need three abilities. They must be able to chase without souring a relationship, because they will be chasing the HIS vendor, the website vendor, the IT team and twelve unit marketing managers in the same week. They must be able to hold state in a document rather than in their head. And they must be comfortable being unglamorous for two quarters.

What they should not be is a brand person with a strong aesthetic point of view. There will be a moment, usually in month two, when someone proposes a website redesign as the answer. A digital operations hire will ask what the current site is failing to do. A brand hire will start a moodboard. You need the first question answered for at least a year before the second one is worth asking.

The second hire: someone who owns the enquiry end to end

The most common structural failure I see in hospital groups is that digital owns the click and the contact centre owns the call, and nobody owns the gap. Leads arrive. Leads are dialled. Appointments are booked somewhere. Footfall happens or does not. At the end of the month, digital reports enquiries and operations reports revenue, and the two numbers have never met.

Your second hire owns that whole chain. Not the contact centre headcount — you will rarely get that in year one, and you should not fight for it yet — but the definition, the routing, the disposition codes, the speed-to-first-call, the follow-up discipline and the reconciliation against the hospital information system. This person spends at least one day a week sitting with tele-callers, with a headset on, listening.

That habit is the hire. If the candidate treats sitting in the contact centre as beneath the role, they cannot do this job in a hospital. Almost everything you need to know about why digital spend is not converting is audible in twenty calls.

The third hire: content with a clinical conscience

Medical content is its own discipline and it is where most digital teams in Indian healthcare quietly fail.

You need someone who can get forty minutes with a consultant, ask questions that do not waste those minutes, come back with a page a patient can actually read, and then take the correction without defensiveness when the doctor strikes out half of it. They need enough clinical literacy to know when a claim has drifted — when “minimally invasive” has quietly become “painless”, or when an outcome has acquired a certainty the evidence does not support.

Hire for temperament over portfolio here. A brilliant consumer copywriter who needs creative ownership will be miserable, and will eventually publish something that costs you a clinician relationship. Someone slightly duller who enjoys getting it right will build you an asset base that compounds for years.

Regional language is part of this role from the start, not a later phase. If a meaningful share of your patients in a Tier 2 catchment are searching and speaking in a regional language, English-only content is not a content strategy, it is a content strategy for the minority of your patients who are least price-sensitive.

The generalist versus specialist call

Hire generalists until a single channel can fund a specialist’s full working week. That threshold arrives later than specialists will tell you and earlier than finance will believe.

Two exceptions, both of which I would make sooner next time.

• Analytics and data. Someone whose whole job is making the numbers agree. In a hospital group, this person’s hardest work is not dashboards, it is reconciling what the website thinks happened with what the HIS recorded and what the unit’s MIS reported. Without them, every meeting becomes a debate about whose number is right and no decision gets taken.
• Someone who can read the hospital systems directly. Not an integration engineer — a person who can get a clean extract of appointments by specialty, by unit, by source, without filing a request and waiting eleven days. The team that can query its own data moves at a completely different speed.

Everything else — paid media, social, design, SEO execution — can be a generalist plus an agency for longer than you think.

What to keep in-house and what to give an agency

The test is not cost. The test is whether the work requires institutional memory.

• In-house: anything touching clinicians, anything touching patient data, the enquiry chain, analytics definitions, medical content sign-off, and the relationship with each unit.
• Agency: production volume, media buying execution, creative craft, technical implementation sprints, and anything where the skill is current rather than contextual.
• Never agency: the list of what is true. Doctor credentials, timings, service availability, pricing guidance. The moment an external team owns that, it decays, because they have no reason to know that a consultant has shifted their Tuesday OPD.

One warning from experience. If your agency is also doing your analytics reporting, you do not have analytics. You have marketing from your agency. Separate those two the week you notice it.

Why this is not a consumer-tech digital team

Every person you hire from consumer technology or D2C will arrive with a set of reflexes that are correct in their last job and dangerous in this one.

They will want to ship weekly. You cannot, because a change to a booking flow can put a patient in the wrong queue at seven in the morning, and the cost of that is not a conversion rate, it is a person in distress at a reception desk.

They will want to test aggressive messaging. You cannot, because claims about outcomes are a clinical and regulatory matter, and because your consultants read your advertising.

They will want to optimise for volume. But the unit does not want volume, it wants the right case mix — and in some specialties in some months, more enquiries are actively unhelpful because the consultant has no capacity and the only result is a poor patient experience and an annoyed HOD.

And the product is not yours. The product is a doctor’s calendar, a bed, an operating theatre slot and a diagnostic machine, all controlled by people who do not report to you. A growth lead who has only ever optimised things they controlled will take nine months to internalise this, and some never do.

The profile that survives the environment

Over two organisations, the people who lasted shared a handful of traits that had nothing to do with digital skill.

• They could be in a seven-thirty morning HOD meeting and be useful in it, without a deck.
• They walked the OPD. Actually walked it, weekly, and knew the front-office staff by name.
• They did not need public credit. In a hospital, the clinician gets the credit. That is correct, and it is permanent.
• They could hold a nine-month time horizon without visible frustration.
• They treated a nurse’s or a front-office executive’s observation as data.

The people who left early were usually excellent and were defeated by the same thing: they needed the organisation to move at the speed of their competence. It will not.

Where the team sits, and who they report to

You will be tempted to ask for unit marketing managers to report into you. Do not ask for it in year one.

I have tried both arrangements. Solid reporting into the group function gives you control and costs you goodwill you cannot rebuild quickly. A dotted line, combined with an explicit service promise from your team — turnaround times, a named person per unit, a weekly call that you never cancel — buys you compliance without a political fight. Earn the solid line by being useful for four quarters, then take it when a unit head asks for it.

The test of whether this is working is simple. When a unit has a problem at nine at night, do they call your team, or do they call a vendor directly? If it is the vendor, you are still overhead.

If you’re building this next quarter

1. Spend three weeks auditing before you write a single job description. Count phone numbers, listings, doctor profiles, logins and live campaigns.
2. Hire the digital operations person first. Give them the audit as their mandate.
3. Borrow, do not hire, for media and design in the first two quarters. An agency on a short contract is a cheaper way to learn what you need.
4. Hire the enquiry owner second, and put the contact centre relationship in writing before they join.
5. Hire medical content third, with a named clinical sign-off process agreed in advance — otherwise they will be blocked in week two.
6. Add analytics as soon as you have two recurring arguments about numbers. You will have them by month four.
7. Name a single point of contact per unit, publish turnaround commitments, and hold to them for a quarter before you ask the units for anything.
8. Only then draw the org chart. By then it will describe something real.

Every hospital digital team I have seen fail was staffed with people who were good at digital. The ones that worked were staffed with people who were good at hospitals and competent at digital.

Questions people ask