He wasn't being dismissive. He was describing a real pattern. The portfolios that win awards on Dribbble and the portfolios that get people hired at startups are optimized for completely different audiences. And most candidates are optimizing for the wrong one.
The 8-second test
I talked to 12 hiring managers at startups ranging from seed stage to Series D. Engineers, designers, PMs, data scientists — different roles, but a remarkably consistent pattern in how they evaluate portfolios and work samples.
Every single one of them described some version of the 8-second test. They open your portfolio or GitHub or project page, and within 8 seconds they've decided whether to keep reading or close the tab. That decision is based on exactly one thing: can they immediately understand what you built and why it mattered?
Not how you built it. Not your design process. Not your tech stack. What you built, and why it mattered.
The portfolios that fail the 8-second test almost always fail for the same reason: they lead with process instead of outcomes. "I conducted 15 user interviews and created an affinity diagram" tells a hiring manager nothing about whether you're good at your job. "I redesigned the checkout flow and conversion went from 2.1% to 3.8%" tells them everything.
What actually works, by role
For engineers
The best engineering portfolios aren't portfolios at all. They're GitHub profiles with 2-3 projects that demonstrate real problem-solving.
What works: A project README that starts with "This solves X problem" followed by a live demo link. Clean code that shows you understand architecture, not just syntax. Contributions to open-source projects that show you can work in existing codebases.
What doesn't work: 47 tutorial projects. A personal website built with a template. A GitHub profile where every repo was last updated 18 months ago.
One hiring manager put it bluntly: "I'd rather see one project that solves a real problem messily than ten projects that implement tutorials cleanly. The first shows me you can think. The second shows me you can follow instructions."
The specific thing that makes engineering hiring managers click "schedule interview": a project where you made non-obvious technical decisions and can explain why. "I used SQLite instead of Postgres because the dataset fits in memory and I wanted zero-config deployment" shows more engineering judgment than "I built a CRUD app with React and Node."
For designers
Lead with the metric, not the mockup. The single most effective portfolio structure I've seen is:
- One sentence describing the business problem
- One screenshot of the solution
- One metric showing the impact
- Then — and only then — the process details for people who want to dig deeper
A product design lead at a health-tech startup told me: "I've hired 9 designers in the last two years. The ones I hired fastest all had the same thing in common — they could connect their design decisions to business outcomes. Not in a hand-wavy way. In a 'this specific change drove this specific number' way."
What kills designer portfolios: case studies that are 4,000 words long with no clear outcome. Dribbble shots that look beautiful but don't show real product work. "Redesigning Spotify's homepage" as a portfolio piece — it tells me you can make things pretty, not that you can ship things that work.
For PMs and data people
The best PM portfolios are blog posts. Not case studies — blog posts. Write about a product decision you made, why you made it, what data informed it, and what happened. This format works because it demonstrates the actual skill: structured thinking about ambiguous problems.
For data scientists and analysts, the equivalent is a well-documented analysis. Not a Jupyter notebook with 200 cells — a clean writeup that starts with a question, walks through the methodology, and ends with a recommendation. Kaggle competition rankings impress other data scientists. A clear analysis that changed a business decision impresses hiring managers.
The three things every portfolio needs
Across all 12 conversations, three elements came up repeatedly:
1. Proof that you shipped something real. Not a concept. Not a prototype. Something that real users used. It doesn't have to be at a famous company — a side project with 50 users counts. The point is demonstrating that you can take something from idea to reality, which is a fundamentally different skill than making mockups or writing code in isolation.
2. Evidence of judgment, not just execution. Anyone can follow a tutorial or implement a spec. What hiring managers are looking for is evidence that you can make good decisions when the answer isn't obvious. This shows up in how you describe tradeoffs: "I chose X over Y because of Z constraint" is the most powerful sentence in any portfolio.
3. Writing that respects the reader's time. Every hiring manager I talked to mentioned this. They're reviewing 20-50 portfolios per open role. The ones that get attention are the ones that communicate efficiently. Short paragraphs. Clear headings. The important stuff first. If your case study requires 15 minutes to understand what you did, most people will never find out.
The anti-portfolio strategy
Here's something counterintuitive: for some roles, the best portfolio strategy is not having a traditional portfolio at all.
Several engineering managers told me they're more impressed by a well-written README on a single GitHub project than by a polished portfolio website. A data science lead said she's hired three people based primarily on their blog posts about technical problems — no formal portfolio at all.
The underlying principle is that the format matters less than the content. A beautifully designed portfolio website with weak projects loses to an ugly GitHub profile with one genuinely impressive project. Every time.
If you're spending more time on your portfolio's visual design than on the quality of the work it showcases, you've got the ratio backwards.
The update that matters most
One last thing that came up in almost every conversation: recency. A portfolio that hasn't been updated in a year signals that you've either stopped learning or stopped caring. Neither is a good look.
You don't need to add new projects constantly. But updating your most recent project, refreshing your bio, and making sure your links work — that basic maintenance signals that you're active and engaged. One hiring manager told me she automatically skips portfolios where the most recent project is more than 8 months old.
The fix is simple: pick your best project from the last 6 months, write it up with the outcome-first structure, and put it at the top. That single update does more for your job search than redesigning your entire portfolio.
