Most portfolios are decks pretending to be websites. They have a hero image, a grid of project thumbnails, and a long PDF you have to download to learn anything substantive. The information architecture says this is a one-way artefact, here to be skimmed.
I don't think that's enough anymore. Recruiters and hiring managers in 2026 are looking at hundreds of portfolios. The ones that get remembered are the ones that behave like products — they have a clear job to do for the visitor, they reward exploration, they have texture, they ship.
The product brief I gave myself
When I rebuilt this site, I wrote it down like a real brief:
- Audience: a hiring manager in Italy with seven minutes between meetings.
- Job to be done: decide whether I'm worth a 30-minute first call.
- Failure mode: they bounce in 12 seconds because the homepage is generic.
- Non-goals: impressing other designers, winning awards, looking trendy.
Every decision after that hung off those four points.
What that produced
A few non-obvious things:
Live demos beat case studies. Every interactive piece on this site — Agora, the hand-tracked cursor, the colour-physics toy — was a deliberate choice over writing another case-study deck. A recruiter clicks once, has a working thing in their browser, and instantly knows I can ship working software. No image, no PDF, no demo video can do that.
EN/IT toggle as proof, not feature. I'm applying for jobs in Italy. Saying "I'm learning Italian" in a CV is cheap. Building a working bilingual site that defaults to English but flips to Italian on a click is the same statement, but it's already shipped.
The CV is downloadable as ATS-clean PDF, not pretty PDF. Most designer CVs are gorgeous documents that lose 40% of their content when an applicant tracking system parses them. I made the version on this site ugly on purpose — single column, plain typography, every keyword machine-readable. The pretty version lives on the website itself.
Brutal pruning. I cut three full pages of "process" content from the about section because nobody who matters cares about my Figma library structure. The pages that survived are the ones a busy reader would actually use.
The lesson
If you're a designer or marketer building a portfolio in 2026, ask yourself one question: if this had to acquire its first 100 users with no advertising, would it? Most portfolios couldn't. The answer to that question is the entire brief.
The portfolio you're reading right now is the answer I came up with. I'm interested in whether it works.