Someone is going to search your name before they meet you. A recruiter, a ministry official reviewing a shortlist, a potential collaborator who half remembers a talk. What they find decides the meeting before it happens.
For most professionals what they find is a LinkedIn profile and possibly a conference bio from 2019. That is not nothing. But it is a claim, and it sits on land you rent.
The difference between a claim and evidence
A CV says you led a multi-country program. Anyone can type that.
A portfolio can show the artefact: the diagram of the reporting hierarchy you designed, the before and after of a data quality curve, the actual feature that shipped, the write-up of how you decided. The claim becomes checkable.
That matters most in specialised work, where the people qualified to evaluate you are few and are not reading a bullet list. They want to see how you think. A well-written case study does more than a decade of job titles.
What you own versus what you borrow
Everything on a platform is subject to that platform. Reach gets throttled, layouts change, features vanish, accounts get restricted by automated systems with no appeal you can reach. None of that is hostile, it is just what renting is.
A domain and a handful of files are yours. Move hosts, redesign, take it down, keep it for twenty years. Your email address on your own domain outlives every employer.
This is the same argument I make to ministries about health systems: own the thing your work depends on. It applies to your own career too.
Being findable
Search is the practical part.
A profile on a large platform competes with millions of pages on that platform. A focused site about a narrow subject can rank for that subject, because almost nobody is writing about it seriously.
Nobody wins on a broad term. But "DHIS2 supportive supervision configuration" or "hospital data migration checklist Kenya" is winnable, and the person searching it has a real problem and possibly a budget. Long-tail search sends fewer visitors and far better ones.
What to actually put on it
A portfolio is not a design exercise. It is an argument, and it needs:
- A clear statement of what you do, in the first sentence. Not a metaphor.
- Three to six pieces of real work, each with the problem, what you did, and what changed. Include the constraint you worked under; that is what tells a reader you have done this before.
- Numbers you can defend. One verifiable figure beats five vague ones.
- Writing. Two or three genuine pieces on your subject do more for credibility than any amount of visual polish, and they are what search engines have to work with.
- One obvious way to contact you.
- A CV people can take away.
What to leave out
- Skill bars. Nobody believes you are 87% at anything.
- Stock photography of people in a boardroom.
- A greeting animation between the visitor and the content.
- Long autobiography. Two paragraphs of who you are, then evidence.
- Anything confidential. In health work especially: describe the method, not the client's data. Where a project is sensitive, write about the approach and leave the specifics out.
On honesty
The temptation is to inflate. Resist it, because in a small field you will meet the people who were actually in the room.
Say what you contributed and what the team contributed. Where a certificate is a course completion rather than an issued credential, write that. It costs you nothing with serious readers and it protects you completely.
It does not need to be complicated
This site is hand-written HTML, CSS and a small amount of PHP. No framework, no page builder, no monthly fee beyond hosting and a domain. It loads fast, works without JavaScript, and I can change any word in it in under a minute.
If you do not build, a template on any reputable host is entirely fine. The technology is not the point. Having somewhere of your own, with real work on it, is the point.
Start with one page and two projects. Add a piece of writing when you have something to say. The version that exists beats the ambitious one you never launch.