This is Clinical Product Thinking đ§ , a weekly newsletter featuring practical tips, frameworks and strategies from the front line of clinical product.
Welcome, friends. This is issue No. 65 of Clinical Product Thinking. This week, weâre answering the question every clinical product manager hears in planning.
In February, I wrote about measured speed: going slow where it matters so you can go fast everywhere else. This issue is the practical follow-up. Itâs about how to answer the question when it lands in your planning meeting.
Because the right answer is: sometimes, yes.
Why ânoâ is the wrong default
Iteration is how good products get made. Real patients and real clinicians find things no test plan will, and a product that never ships never helps anyone.
If clinical product answers every request to ship with âitâs not safe yetâ, two things happen. The team stops asking, and the clinician becomes the blocker everyone works around. Neither makes patients safer.
The useful answer is specific. Some things can ship today and improve next month. Some should ship as a smaller version. A few need to be right before anyone sees them.
The skill is knowing which is which, and being able to explain why.
Five questions to ask before you ship and iterate
1. Could a patient come to harm before the next iteration?
Picture the version youâre about to ship working exactly as built, with all its gaps, for the weeks until the fix arrives. Whatâs the worst that happens to a patient in that time, and how likely is it?
A confusing onboarding screen costs you a few sign-ups. A symptom checker that misses a red flag for six weeks might have an irreversible cost.
2. Will you know if it goes wrong?
Iterating later only works if you find out what needs fixing. Some failures are loud: complaints, support tickets, an obvious drop in a metric. Others are silent. A patient who was wrongly reassured doesnât get in touch to tell you. They just donât come back, or they turn up somewhere else, sicker.
If a failure would be silent, âiterate laterâ could mean never as other things get prioritised. Consider the monitoring first, or not shipping that part yet.
3. Can you undo it?
Jeff Bezos described decisions as one-way and two-way doors. A two-way door can be walked back through. A one-way door canât.
Copy, layout and most interface choices are two-way doors. Clinical logic that other features come to depend on, a data model that shapes what you can record, advice that patients have already acted on, and the expectations you set for clinicians are much harder to reverse. Once a patient has stopped a medication because of your product, the next release canât undo that.
4. Does it change what the product is?
A feature can quietly change your intended use. Adding a score that clinicians use to adjust treatment, or advice that steers a patient towards a clinical action, can move a product into medical device territory one release at a time.
You canât iterate your way out of that after launch. Ask the question before the feature is built: could this drive a clinical action? When âSupportâ Becomes âTreatmentâ walks through the most common grey zones.
5. Who covers the gap until the next release?
Every âweâll fix it laterâ lands on someone. Usually itâs a clinician reviewing cases by hand, or an operations team running a workaround.
That can be a perfectly sensible plan. However it needs to be a deliberate one, with the people covering the gap knowing theyâre doing it, and for how long.
Three answers to give
Once youâve run through the questions, most requests fall into one of three groups.
Ship it and iterate. Low potential for harm, failures youâd spot quickly, easy to reverse. Most interface changes, copy, non-clinical features and internal tools sit here. Go fast, and donât make the team wait for a full clinical review.
Ship a smaller version. The full feature carries real clinical risk, and part of it carries very little. At Ada Health, Lorenzo Espinosa described how teams changed the question. They stopped asking whether a feature could ship, and asked what a first release needed to include to be clinically viable. For an EHR integration, that meant ingesting the minimum clinically necessary data first, as defined by clinical product, then expanding based on real-world use.
Other ways to shrink the risk: release to a small group of patients first, put a clinician review step on the first set of cases, or switch the feature on for lower-risk pathways before higher-risk ones.
Get it right before it ships. High potential for harm, failures that would be silent, or decisions you canât take back. Triage and escalation logic, red-flag detection, dosing, anything that changes your intended use. These deserve the slower, more careful version, and the team should hear clearly why.
Make âlaterâ real
âIterate laterâ only works when later actually happens. Before you agree to it, write down four things:
What will be improved, specifically
When, as a date or a sprint
Who owns it
What signal will tell you it needs to happen sooner
Then put it where the team will see it: ideally on the roadmap, and in the hazard log if thereâs a clinical risk attached.
What to say in the meeting
Next time someone asks âCan we just ship this and iterate later?â, try answering in three parts:
âYes for the new onboarding flow. A smaller version of the reminders, starting with the lower-risk pathway. And the escalation logic needs to be right first, because if it fails we wonât find out until a patient comes to harm.â
Itâs a yes, a smaller yes and a clear explanation. Thatâs harder to argue with, and it shows the team that clinical product is there to help them ship.
The takeaway
Most of what your team builds can ship and improve later. Your job is to spot the few things that canât, say so early, and give a reason the team can act on. Do that consistently and âCan we just ship this?â stops being a standoff in planning. Clinical product becomes the reason the team can move at safe speed.
Help shape the State of Clinical Product 2026 đ
Every December, I publish the State of Clinical Product, my annual look at how our discipline is changing. Last yearâs report drew on more than 100 conversations. This year, I would love to hear from you.
It takes two minutes, answers are anonymous, and the survey closes on Sunday 6 December. Leave your email to get the report early and enter the draw for one of two 1:1 strategy calls with me.
đ Take the survey
If youâre trying to move into clinical product, I especially want to hear from you. Thereâs a short section just for you.
Thatâs all for this week. See you next time! đ
đ¤ Work with me | đ Attend an event | âď¸ Send a message
Written by Dr Louise Rix, Clinical Product, AI, doctor and ex-VC. Passionate about all things healthcare, healthtech and clinical product (âŚobviously). Based in London. You can find me on LinkedIn & Instagram.
Made with đ for better, safer HealthTech.


