Mystery patient programmes: what they measure and what they find
How to run a mystery patient programme without overspending, exactly what to evaluate, and why the problem is almost always the phone.
Digital Marketing, Medical Marketing & AI Consultant · btodigital
Most practices I audit lose patients before the consultation, not during it. They lose them on the phone nobody answers, on the WhatsApp replied to the next day, and with the person who cannot say what it costs.
A mystery patient programme is simply measuring that with someone posing as a patient. You do not need to hire a firm to start.
What exactly gets evaluated?
Not “quality of care”, which is impossible to grade. Concrete, timeable things:
| Moment | What gets measured |
|---|---|
| Phone call | How many rings, whether they answer, whether they know the price |
| WhatsApp message | Minutes until first reply |
| An awkward question | Whether they can say what is and is not included |
| Booking | How many steps and how much information they ask for |
| Confirmation | Whether it arrives, and when |
| Arrival at the practice | Actual wait versus promised wait |
| Departure | Whether the next step is explained |
What makes the exercise useful is that these are times and facts, not impressions. “They were nice” is useless. “Answered on the fourth ring and did not know the follow-up price” is not.
How do you run it without overspending?
You can start with three calls and three messages, made by somebody the team will not recognise:
- Define what you will measure before calling, with a list of five to seven points.
- Have an outsider make the calls. A trusted relative works. It cannot be a team member.
- At different times, including one out of hours and one at peak.
- Record exact times, not feelings.
- Repeat in three months to see whether the change held.
A specialist firm adds value once you have several locations or need the result to be indisputable to the team.
What do you almost always find?
With very few exceptions, this:
- Nobody is certain about the price, or each person says something different.
- WhatsApp is answered late and without a personalised greeting.
- Nothing exists out of hours, not even an automatic message.
- Promised and actual waiting times do not match.
- Nobody explains the next step on the way out.
None of those five requires marketing investment. All of them cost patients.
How do findings turn into change?
This is where most programmes stall. The report arrives, gets read, and nothing happens.
What works:
- One change at a time, with an owner and a date.
- Present times, not judgements. “We took 4 hours to reply” attacks nobody; “we give bad service” does.
- Measure again in three months and show the team the difference.
- Acknowledge what went well. If the exercise only produces reprimands, the next round gets sabotaged.
Is it ethical to do this without telling the team?
A fair question, and the practical answer is to announce that service measurements happen periodically, without saying when. The team knows the programme exists, does not feel spied on, and the measurement stays valid.
What you should not do is use the result to discipline an individual. The target is the process, not the culprit.
Frequently asked questions
What is a mystery patient programme in a medical practice? Someone poses as a patient and times every step: the call, the message, the booking, the wait and the departure. It reveals what the team cannot see from inside.
Do I need to hire a specialist firm? Not to start. Three calls and three messages made by an outsider at different times already surface the main findings. A firm is justified with several locations.
What should be measured? Times and facts: how many rings, how many minutes to reply, whether they know the price, how many steps to book, actual versus promised wait. Never general impressions.
Which problem appears most often? That nobody is certain about the price, and that WhatsApp is answered late. Both cost patients and neither requires advertising spend.
How often should it be repeated? Every three or four months. Less often will not show whether the change held; more often wears people down.
Should the team be told? It is worth announcing that periodic service measurements happen, without saying when. That way nobody feels spied on and the measurement stays valid.
Does it work for small practices? Even more so, because in a small practice one person answers the phone and everything depends on them. The finding is immediate and so is the fix.
