I redid my portfolio over the last few months. Three years of work, nothing older. Floor at the back end, the Tortuga apps in the middle, Crew at the front. Everything else came down.
The first cut was easy. Anything older than three years went, including a few projects I'd been holding onto for almost a decade. Not because the work was bad. Because the kind of work I'm doing now barely resembles what I was doing then. I'm writing more code, shipping more of my own things end to end, making product and growth decisions on the same projects I'm designing. A portfolio of pretty Figma files from 2020 doesn't show any of that.
The second change was harder. I stopped cropping the projects down to just the design work. Josh and I shipped six apps in 2025 under Tortuga Labs. I designed them, wrote a real share of the code, made the marketing sites, picked the pricing, talked to the early users. The Figma frames are the smallest part of what I have to show. If I'd cropped to those, I'd have been hiding the job.
Code stopped being the bottleneck. So the work moved - up the funnel into product thinking, and down the funnel into distribution and growth. A modern portfolio has to show those moves, not just the middle.
The designers I kept open in tabs while rebuilding mine are doing exactly this:
The work and the writing are weighted about evenly on all three sites. The site is a body of evidence, not a gallery.
I'll only ever spend 60 seconds on a portfolio before I decide whether to take the meeting. In that 60 seconds I want to know two things. Is the bar high enough. Is the work the kind of work my team does. If both answers are yes, you're in the room. That's the whole job of the portfolio.
So the new case study, for me, is a live product I can try. The writing matters more than it did three years ago, partly because AI has made it easier to put pleasant, unobjectionable words on a page, and an AI-shaped portfolio post is one of the fastest ways to make me close the tab.
If you're rebuilding yours, cut everything older than three years and see what's left. If what's left is thin, that's information. Make more, ship more, write more.