Insights · Method
The Engineered Practice: applying DMAIC to a clinic
Mohammed Ismail · 2026-10-03
Most practices try to grow by adding traffic: more ads, more posts, a new agency. Sometimes it works for a while. Then costs rise, patients drop off after the first months, and the owner ends up answering messages at night.
Engineers have a different habit. Before making a process bigger, they make it reliable. The tool they use is called DMAIC.
What DMAIC is
DMAIC is the improvement cycle at the heart of Six Sigma. It has five steps:
| Step | Question it answers |
|---|---|
| Define | What process are we improving, and for whom? |
| Measure | How does it perform today, in numbers? |
| Analyze | Where and why does it fail? |
| Improve | What changes remove those failures? |
| Control | How do we keep the gains from fading? |
Factories use it to cut defects. Hospitals use it to cut waiting times. It works on any process with steps, handoffs and outcomes you can count.
A clinic is a process
From the outside, a practice looks like a set of people and tools. From an engineer's view, it is one flow:
- Someone searches, reads an AI answer or sees an ad.
- They land on a page and decide whether to book.
- They book, or they don't.
- They complete intake, or they stall.
- They show up for the first visit, or they don't.
- They stay past month three, or they quietly leave.
Every arrow in that list is a place where patients leak. Most practices only measure the first step (clicks) and the last one (revenue). The leaks in between stay invisible.
From DMAIC to The Engineered Practice
I adapted DMAIC into five stages that fit how practices actually work:
| DMAIC | The Engineered Practice | What happens |
|---|---|---|
| Define + Measure | Map | Draw the full patient flow and measure each step |
| Analyze + Improve | Fix | Repair the steps that lose patients today |
| Improve | Automate | Remove manual work that slips on busy weeks |
| Improve | Grow | Turn on acquisition once the flow can convert |
| Control | Control | Review the same numbers monthly and keep improving |
The order matters. Growing before fixing means paying to send people into a leaky flow.
The five numbers to start with
You don't need a data team. Start with five numbers:
- Cost per booked visit. Not cost per lead. Leads that never book hide the real price.
- Show rate. How many booked first visits actually happen.
- Lead to visit time. How long from first inquiry to first visit.
- Month-3 retention. How many patients are still active after three months.
- Admin hours. How many hours each provider spends on admin after hours.
When these five are visible every month, most decisions get easier.
Ranking what to fix first
Not every leak is worth fixing first. Engineers rank failures with FMEA (Failure Mode and Effects Analysis). Each leak gets three scores from 1 to 10:
- Severity: how much revenue or trust it costs.
- Occurrence: how often it happens.
- Detection: how hard it is to notice.
Multiply them and you get a Risk Priority Number. The highest numbers go to the top of the list. It is simple, and it stops the loudest opinion in the room from deciding.
Why Control is the stage most practices skip
Improvements fade. A new reminder flow works well for two months, then someone changes a template and show rates slide back. Control is the routine that catches this: a dashboard, a monthly review, and a simple control chart that shows whether a change in the numbers is real or just normal noise.
Where to start
Start with the map. One page, every step from first click to month six, with the tool and the person behind each step. If you want a quick outside view first, the free Readiness Score scores your practice on all five stages in about three minutes.
Key takeaways
- Map the whole patient flow before spending more on traffic
- Measure cost per booked visit, show rate and month-3 retention
- Fix leaks before scaling ads
- Automate anything that depends on someone remembering
- Review the same five numbers every month
General information only, not legal or medical advice.