---
project: Portfolio Website
updated: 2026-01-15
---

# Portfolio Website — Decisions

## 2026-01-05 — Next.js App Router, not Pages Router

**Decision:** Use Next.js App Router (not Pages Router) for the portfolio.

**Why:** App Router is the current direction for Next.js; using it means the codebase won't need a migration later. The portfolio is simple enough that the App Router's complexity doesn't add meaningful friction — it's mostly static content with one contact form.

**Trade-off:** Some ecosystem packages (particularly older auth libraries) still have rough App Router support, but this project doesn't need auth, so that's a non-issue.

---

## 2026-01-08 — No CMS; content in TypeScript constants

**Decision:** Store all portfolio content (project descriptions, case study text, bio) in TypeScript constants files, not a headless CMS.

**Why:** The update frequency is low (a few times a year at most). A CMS adds infrastructure cost and complexity that doesn't pay off at this scale. If the site eventually needs frequent updates or multiple editors, revisit.

**Trade-off:** Updates require a code change and redeploy. Acceptable for a personal portfolio.

---

## 2026-01-10 — No analytics on launch; add after 30 days

**Decision:** Launch without any analytics tool. Add Plausible (privacy-friendly, ~$9/mo) after the portfolio is actually getting traffic worth measuring.

**Why:** Analytics before meaningful traffic is noise that invites premature optimisation. First priority is getting the portfolio live and in front of people.

**Trade-off:** No baseline data from day one. Acceptable.