The doctor-referral network as a product

In every Indian city I have worked in, a large share of a tertiary hospital’s admissions arrive because another doctor sent them. A family physician in a residential colony. A nursing home with twelve beds and no intensive care. A specialist in a district town three hours away who does the diagnosis and sends the surgery to the city. The patient has already been told where to go before they have heard of your brand, and the person who told them is not on your payroll and not in your CRM.

This network is the least digitised demand channel a hospital group has, and the one most encrusted with practices nobody wants to discuss in a board meeting. The growth function usually inherits it as a field team called something like doctor relations, a set of relationships that live in individual phones, and a quiet understanding that some of the volume is bought. What it rarely inherits is data.

I want to treat the referral network the way I would treat any other channel: as a product with users, a journey, a measurement problem and a set of rules. The users are referring doctors. The journey is what happens to their patient and what comes back to them. The measurement problem is real and solvable. The rules are not negotiable, and I will be plain about them.

How it actually works

Strip away the euphemisms and the referral economy in an Indian city runs on three currencies. The first is trust: the referring doctor believes your specialist will treat their patient well and send them back. The second is reciprocity: the referring doctor gets something for the referral, which in the honest version is information, access and respect, and in the dishonest version is money. The third is convenience: it is easy to send the patient to you, the patient gets seen quickly, and the doctor hears what happened without chasing.

Most hospital groups compete on the first currency through their specialists’ reputations, have an uneasy relationship with the second, and ignore the third entirely. The third is where a digital operator can build an advantage that a competitor cannot buy, and it is where I would put the product effort.

The other thing to understand is that the network is personal, not institutional. The referring doctor sends to a surgeon, not to a hospital. When that surgeon moves, the referrals move with him, and the group discovers that a channel it thought it owned was on loan. Part of the work is making the institution a party to the relationship, so that the referring doctor’s experience is with the group and not only with one consultant.

The ethical boundary, before anything else

The group does not pay for referrals. Not as a fee, not as a “professional charge” for a consultation that did not happen, not as a gift, not as a conference sponsorship contingent on volume, not through a liaison officer who is not asked where the money went. This is a legal matter, a regulatory matter and a brand matter, and it is also a growth matter, because a channel built on payment is a channel a competitor can outbid tomorrow.

Saying this in a policy document is easy. Making it true requires the growth function to give the doctor-relations team something else to offer, and to measure them on something other than volume from their list. If a field executive is targeted on admissions from named doctors and given a discretionary budget, you have built the incentive you claim to prohibit. Target them instead on referring doctors onboarded to the portal, reports delivered within the promised time, feedback calls completed and complaints resolved. The volume follows, and it follows for reasons you can defend.

I would also audit the “referred by” field in the hospital system against consultation billing and against the liaison team’s expense claims once a quarter. Not because you expect to find something, but because the people who might be tempted should know that you look.

What the referring doctor actually wants

I have sat with enough family physicians and district specialists to have a consistent list, and it is short. They want their patient seen quickly, because the patient will otherwise come back to them and ask why they were sent somewhere with a three-week wait. They want the specialist to acknowledge the referral, ideally with a call or a message, because it tells them their judgement was respected. They want the discharge summary and the key reports, promptly, so that when the patient returns to them for follow-up they are not working blind. And they want the patient sent back to them for that follow-up rather than absorbed into the hospital’s own outpatient department.

That last one is the one hospital groups get wrong most often, and it is the one that poisons the relationship fastest. A referring doctor who sends a patient and never sees them again has learnt that you take patients rather than share them. The rule that follow-up care goes back to the referrer wherever clinically appropriate is a commercial rule as much as a courtesy, and the product should make it the default path rather than an exception someone has to remember.

Report sharing is the core of the product

If you build one thing for referring doctors, build the way they receive information about their patient. A portal or an app in which a doctor registered with the group can see the status, the discharge summary and the key reports of every patient they have referred, with the patient’s consent captured at registration. A message the day the patient is admitted, the day of the procedure and the day of discharge. The specialist’s note, in a form the referring doctor can act on, not the internal record.

Consent and privacy are the constraints. The patient decides whether the referring doctor sees their information, and the product must record that decision and honour it. In practice, patients almost always say yes, because they want their family doctor involved. The point is that the system asks.

Most groups do this by messaging app today, from the consultant’s personal phone, with photographs of reports. It works until it does not: a consultant leaves, a phone is lost, a report goes to the wrong doctor, a patient discovers their records were shared without being asked. Moving this to a product the group operates is a compliance improvement and a relationship improvement at once, and it is the first step to the institution being a party to the referral.

Closing the feedback loop

The referral relationship dies in silence. A doctor sends a patient, hears nothing, and quietly sends the next one elsewhere. The product should make silence impossible.

Beyond reports, that means a feedback call or message to the referring doctor after each episode: was the patient seen on time, did the specialist speak to you, did the summary arrive, is there anything the group did wrong. It means a visible path for the referring doctor to raise a problem, with someone who answers. And it means the specialist knowing which doctors referred to them last month, so that they can make the call that keeps the relationship alive. A consultant who is shown a list of their top referrers and asked whether they have spoken to each of them this quarter usually has not, and usually will.

The other half of the loop runs toward the patient. A patient who has been referred and does not book, or books and does not come, should trigger a follow-up from the contact centre, because the referring doctor will hear about it before you do. The CRM should know the referral came from a doctor, and treat it as an enquiry that already has a reason and a recommendation attached, because those convert at a rate no campaign matches.

Measuring referral share without paying for it

You cannot manage this channel if you cannot see it, and most groups cannot. The registration desk asks “who referred you?” and types whatever the patient says, which is often the wrong name, a clinic name, or “self” because the question was not understood. The referral field is unusable for analysis and everybody knows it.

Fixing it is a data and a product task. A maintained directory of referring doctors, with a unique identifier, that the registration desk searches rather than types into. The referral question asked in the contact-centre script and the booking flow, not only at the desk. The referring doctor’s identifier carried from enquiry through appointment to admission, so that the referral funnel can be seen whole. And a monthly view, by unit and speciality, of admissions by referral source: self, doctor-referred, insurer, corporate, scheme, with the doctor-referred segment broken down by referrer and geography.

Once that exists, you can see things that used to be argued about. Which specialists actually draw referrals and which are living on walk-ins. Which districts send patients and which do not, and whether that matches where the liaison team spends its time. Which referring doctors have gone quiet in the last quarter, which is the list the specialist should be calling. And, over a year, whether the portal, the reports and the feedback calls moved referral volume, without a rupee changing hands. That is the evidence I took into the executive committee to fund this as a product rather than as a field team, and it held.

Where the field team fits

None of this removes the need for people who visit doctors. It changes what they do. The doctor-relations executive becomes the person who onboards the referring doctor to the portal, shows them how their patients’ reports arrive, brings the specialist for a visit, carries back complaints and makes sure they are resolved. They stop being the courier of favours and become the account manager of a relationship the institution now has data on.

That transition is uncomfortable for some of them, and a few will leave. In my experience the ones who stay are relieved, because they were never comfortable with the old currency either.

If you’re starting this next quarter

• Weeks one to three: pull whatever referral data exists and admit how bad it is. Build the referring-doctor directory from consultant lists, the liaison team’s contacts and the registration field, and deduplicate it.
• Weeks three to six: change the registration and contact-centre flows to search the directory. Add the referrer identifier to the CRM record. Capture patient consent for sharing with the referrer.
• Weeks six to ten: launch report delivery to referring doctors from a group-operated channel, starting with the two or three specialities that carry the most referred volume. Set a promised turnaround and measure it.
• Weeks ten to twelve: rewrite the doctor-relations team’s targets around onboarding, turnaround and feedback. Give each specialist their referrer list. Set up the quarterly audit.
• Next quarter: present admissions by referral source to the executive committee and the medical director, and let the product decisions follow the data.

A referral you paid for belongs to whoever pays more next year. A referral you earned is yours to lose, and you will know why.

Questions people ask