A fixed-fee medical consultation contains a hidden adversarial game.

Once the fee is set, the doctor earns the same amount whether the consultation takes five minutes or 20, so shorter is better for them. The patient has already paid (or implicitly agreed to pay), so more time feels like better value. The predictable strategies follow: doctors try to keep appointments short, while patients bring lists of extra concerns to get their money’s worth. The fixed fee puts them on opposite sides of the clock.

Clinic infrastructure reinforces this equilibrium. Premises, staff, Medicare item numbers, billing software and patient expectations are all built around appointments with a fixed price and duration. Changing the invoice while leaving the rest intact would create more friction than it removes. A workable per-minute clinic has to be designed around the meter from the start: video only, fully private, no premises or staff, with intake, scheduling and billing automated where possible.

Bill by the minute

Per-minute billing gives the patient a natural off-switch. They stay while the next minute is worth more to them than it costs, then leave when it is not. The doctor is paid for the time used, so duration no longer creates the same conflict.

The meter also exposes what patients value inside a consultation. An appointment mixes medical information with reassurance and the rituals through which a doctor is seen to be diligent and caring. A fixed fee makes the marginal cost of those extra minutes zero once the appointment starts; a meter makes the cost visible. Patients can then decide which parts are worth buying. That choice may reveal how much of a conventional consultation was medicine and how much was signalling.

What changes downstream

Preparation now has a direct return for the patient. Someone who arrives with “I have this thing, it’s been there a while, I tried some cream” might spend five paid minutes transmitting information they could have organised beforehand; clear photographs and a structured history might reduce a 12-minute consultation to three. The time and money are the patient’s, so the incentive to prepare sits with them.

The same incentive gives AI preparation a concrete job: an LLM can turn a disorganised history into a structured summary and shorten the paid consultation. If the clinic supplies the tool, it can control the prompts and output format, although how far a medical service should embed AI in intake remains an open question.

Per-minute billing also changes the doctor’s response to disorganised use of time. If a patient wants the doctor to spend 10 minutes reading biopsy reports they did not summarise, that choice is explicit and paid for, so there is less reason for quiet frustration or for policing how the patient uses the appointment.

The alignment remains incomplete because the doctor controls more of the pace and has more information about what the consultation requires. Since a patient may not detect padded time, reputation is the main counterweight: a doctor who consistently takes longer than peers should lose patients. The meter makes the asymmetry visible without removing it.

A prefunded account, perhaps a $500 deposit at signup, removes payment chasing and the awkward bill at the end of each call. The clinic stays fully private, without Medicare billing, bulk billing or rebates. This narrows the patient pool to people willing to pay out of pocket, while removing a large layer of administration. In return, the transaction becomes direct: the patient pays the doctor for time.

A clinic without staff

An AI receptionist handles calls and routine patient communication; an AI scribe produces notes and referral letters from consultation audio; ordinary software handles scheduling, metering and payment. With video consultations and no premises, one specialist can run the operation.

The platform underneath

The same components work across specialties: scheduling, per-minute billing, AI intake, AI scribing, reception, payments and video with a visible timer. Any specialist could run a clinic on the same stack.

The simplest commercial route is to run one clinic and prove the model. The platform could instead become clinic-in-a-box software for other doctors, or remain private infrastructure used to recruit specialists. In the latter model, the business captures a share of consultation revenue instead of subscription fees.

Technical constraints

  • Prefer files to apps, so data remains portable and every component can be replaced.
  • Use Stripe’s usage-based billing because metering and charging are core infrastructure.
  • Keep control of the data, with encrypted backups, verified restores and Australian residency for patient records.
  • Self-host patient records in Australia, while low-consequence components such as the marketing site run on ordinary cloud infrastructure.
  • Custom-build the patient experience and billing engine, and use the simplest workable option everywhere else.
  • Keep the whole system buildable and maintainable by one person; an architecture that requires a team defeats the premise.

The bet

The bet is that the unit of billing determines the rest of the clinic. Give the patient an off-switch and pay the doctor for time, and the leaner operating model follows from the incentives.

Running one clinic is a solo project; turning it into a platform or specialist network probably needs a co-founder. If you want to build that, get in touch.