What pharmacy taught me about product decisions

I trained as a pharmacist while already working in product. During my internship year at Lagos University Teaching Hospital, I ran the hospital's antimicrobial stewardship programme, and worked at a product studio at the same time. For a year I moved between a ward and a backlog.

The stewardship programme shaped how I approach product decisions. Four habits from that work still guide my research, prioritisation and delivery.

The programme

Antimicrobial stewardship is the work of making sure antibiotics are used appropriately. Overuse and misuse drive resistance, so hospitals run programmes to change how antibiotics are prescribed. I ran the design, rollout and continuous improvement of one across a large teaching hospital.

It is a hard change to make. The people whose behaviour has to change are senior clinicians making judgements about their own patients. The person asking them to change, in this case, was an intern pharmacist.

1. Evidence before opinion

I analysed 36 months of prescribing data and turned the findings into policy. Inappropriate antibiotic use fell by 40%.

As a junior pharmacist, I needed a case clinicians could assess. Three years of their hospital’s prescribing data gave us a concrete basis for discussing change.

I carry this directly into product work. I define the metric an initiative will be judged against before it ships. I set a baseline before changing anything. And when there is a disagreement about priorities, I would sooner bring a query result than an argument.

Three years of a hospital's own data is hard to dismiss as a bad month.

2. Influence without authority

None of the physicians, pharmacists, microbiologists or hospital leaders reported to me. I could not instruct anyone. The programme worked because those groups agreed on shared priorities, a reporting rhythm and common standards, and I coordinated them around those.

Product managers often face a similar situation: the people needed to deliver a change do not report to them. I work towards agreement by explaining the problem in terms each group cares about, sharing the evidence and making the next step clear.

I used the same approach later in a private hospital, where five departments had competing priorities. One shared, visible backlog did there what shared reporting did in the stewardship programme.

3. Adapt the standard to the place

I also co-authored the protocol for the hospital's first bone-marrow transplant programme. International standards for this exist. The work was adapting them into a workflow that this team, in this hospital, could run.

Product managers face the same task with best practice. A framework written elsewhere assumes someone else's team, tools and customers. Copying it whole rarely works. The useful question is which parts of the standard are essential for safety or quality, and which are assumptions about resources that do not hold here.

4. Adoption, not launch

A policy that is written and not followed changes nothing. The stewardship programme needed reporting frameworks, operating documents and staff training before it became part of daily practice.

I bring that view to product delivery. Launch is followed by the work of adoption: training, documentation and follow-up. I assess whether people use the product and whether the intended outcome improves.

Research habits

Before the internship I worked as a research assistant in clinical pharmacy. I ran mixed-methods research with 290 participants, combining structured surveys with in-depth interviews, and contributed to a peer-reviewed publication.

Two habits stay with me: combining quantitative patterns with the context interviews provide, and keeping fieldwork consistent enough to compare findings. Both inform how I run product research.

Why it matters for regulated products

Much of my product work has been in regulated or high-trust settings: hospital workflows, payments, and an AI assistant for lawyers. Pharmacy is training for exactly that.

• A pharmacist checks before dispensing. I build the check into the workflow before release.
• A wrong dose is not a bug to fix next sprint. I specify the failure paths first.
• Every decision in a clinical record can be traced. I expect the same of a product that makes decisions for its users.
• Clinical safety, compliance and usability pull against each other. Balancing them is the job, in a hospital and in a product.

I do not think of pharmacy and product as two careers. One is where I learned to decide carefully. The other is where I do it.