What’s on the bench

The site’s structure: how four different sides of my life share one domain without turning into a junk drawer.

The decision

One site, path-based sections, no subdomains. A hub homepage; four pillars — legal tech, running, the workshop, and field notes; a speaking page under legal tech; and a single newsletter that feeds everything.

How I took it apart

My first instinct was subdomains — run.zackglaser.com, build.zackglaser.com — a tidy brand for each version of me. Three things killed it:

Subdomains split search authority. Search engines treat each one as roughly its own site, so four subdomains means four weak sites instead of one that compounds.

Subdomains quadruple maintenance. Four designs, four deploys, four things to feel guilty about neglecting. I know my failure mode.

Subdomains contradict the premise. If the brand is “the mix is the point,” then walling the audiences off from each other is architecture arguing with strategy.

So: sections under one roof, built as tagged collections. Every post carries a pillar tag; the section pages are just filtered views. The escape hatch is that this makes any pillar promotable later — if the running content ever earns its own home, pointing a subdomain at an existing collection is a weekend job, not a rebuild.

The other structural bet is the newsletter as the engine. Not four content streams — one format that rotates across pillars: each issue takes one thing apart. A legal tech tool one week, a marathon training block another, a workbench build, an idea. One publishing habit feeds the whole site, and each issue auto-files into its pillar’s archive.

Speaking gets a dedicated page because it’s the site’s one conversion job — topics, past appearances, one-sheet, booking form. Everything else builds trust; that page closes.

What I got wrong (or don’t know yet)

Two unknowns. First, section names: “field notes” for the essay pillar feels right today; “the workshop” might collide with how I describe physical projects versus code. Naming needs its own entry. Second, I haven’t validated whether one rotating newsletter retains subscribers who came for a single pillar — the crossover is the bet, but it’s still a bet.

Next up

Where the code actually lives, before any framework gets picked.

Costs so far: $12/yr / 3 hours