Back to essays

Why I'm writing my portfolio in code, not Behance

April 26, 2026 · 6 min

The pattern I see most often on designer portfolios is the Behance approach: design every case study as a series of high-resolution images in Figma, export as PNG, upload. Pixel-perfect control. Beautiful on a 1440px display. The problem starts when someone opens it on a phone.

A 1920px-wide image scaled to fit a 375px iPhone screen renders 12pt body text at about 2.3pt. Readers either pinch-to-zoom every section or bounce. Behance hides this from designers because most viewing happens on desktop, and the platform handles its own light-touch responsive scaling. On a personal site, the same approach falls apart.

There are real costs to going the Behance route on your own domain:

  • Mobile is unusable. Recruiters open links from LinkedIn on their phones. The "responsive" version of a 5MB rendered slide is a microscopic version of the same thing.
  • Text isn't text. Screen readers, search engines, copy-paste, and "Find on page" all see nothing. A recruiter searching "ethnographic" can't find your Knot case study.
  • Performance. A page with five 1920×4000px PNGs is 5–10MB. Slow on mobile data, hurts Core Web Vitals.
  • Maintenance. Fix a typo? Re-export from Figma, replace the file. Iteration cost goes up 10×.

The deeper objection is the one designers tend to push back on: isn't designing in code a constraint on creativity? I don't think it is anymore. Modern CSS is more flexible than Figma in many ways — animation, scroll-triggered reveals, type that adapts to viewport, and now custom properties that let a single component look completely different on different pages.

The 'rendered images' approach is a workaround for a problem that doesn't really exist anymore.
Me, after spending a week building this site

What I've ended up with is a hybrid. Real text where text should be — the case study narrative, the reflections, the methodology. Strategically placed images for the visual artifacts of the work — UI mockups, ethnography photos, the actual deck slides for diagrams that aren't worth rebuilding in CSS. The case study reads on a phone. The diagrams open into a fullscreen viewer when you want detail.

The trade I made was: more upfront engineering work, in exchange for content that's accessible, fast, and editable in 30 seconds. Six months in, I think it's the better trade.