1. Agencies and in-house teams want different proof
First up was a generalist who wants to join an agency. We said up front that pitching yourself as a freelancer or consultant isn’t our lane. Neither of us has ever sold ourselves as a consultant, so we kept it high level and tried not to claim experience we don’t have.
A portfolio has two jobs: introduce who you are and demonstrate value, and value depends on who’s looking. In-house teams tend to look for strategy. Freelance clients want the finished product. Agencies mostly want production — high-quality output, and a lot of it. Tim mapped it as difficulty against ownership: freelance and agency work both come scoped and leave you little ownership, but agencies set a higher quality bar, while in-house is where you get to own strategy and running your own design business can be the hardest of all. It’s a simplified map, and Tim said so, but he thinks it holds in general.
So decide where you’re applying first. For an agency, show lots of projects across lots of clients, a little process, and mostly outcomes and finished output. If what you really want is to own decisions, Tim’s read was that agency work points the other way, and going in-house is the likelier fit.
2. Centralized teams hire the skill they don’t have yet
Nate walked through team formations. In a centralized team, every designer works under one design manager, picks up assigned tickets and hands them off to engineering. When that manager hires, they’re looking at the team they already have. Someone covers accessibility, someone else covers design systems. They don’t need a second accessibility person, so the real question is what you do that nobody on the team does yet.
Agencies usually run centralized, and that rewards specialists. Decentralized setups, more common in enterprise, put one designer on a product team with PMs and engineers, and that person has to be passable at everything. For agency work Nate would rather you be really good at one facet than mediocre at many. Later in the session he made the broader case for a niche: the narrower you go, the fewer people you’re competing with, and getting interviews is the whole point of the portfolio.
Pick the facet and point the whole site at it. Work that pulls elsewhere, like logo design on a UX engineering pitch, comes off the main portfolio. Give it its own site for other applications, or a “play” or “other work” section. Tim was clear this isn’t a judgment on whether the work is good, only on whether it helps the story you’re telling. One quick test from a later review: if the line at the top of your homepage could be pasted onto almost any designer’s site and still be true, it’s too broad.
3. Lead with why the work mattered, then what you built
A mid-level designer was getting callbacks that never turned into first-round interviews. Tim drew the hiring funnel. The ATS only reads your resume, and the recruiter mostly does too, maybe glancing at the portfolio for a vibe check. The HR or hiring manager in the middle looks at both and decides who’s worth the design lead’s time, and the design lead looks at the portfolio. Those people want to know if you think like a designer and understand why your work mattered. You still have to show the work. It just isn’t the proof.
The case study titles named what got built: a dashboard, which is basically a screen. Tim pulled up our metrics-by-value chart. On one side sits output, like screens shipped, flows or a component library. On the other sit engagement, behavior change and time saved, and those mostly roll up to the business making money through conversion or retention. So flip the title. Lead with the impact, say who it was for, then say how. And drop “AI-powered” as a selling point. It’s a method, and you wouldn’t call something a “cloud-hosted dashboard” either.
Tim’s broader point was that designers get trained to produce output instead of explaining it, and explaining your reasoning is what shows you can handle any problem. So open each case study with the full problem in one sentence, without using the word “problem.” Follow it with the move that obviously comes next, then the steps. It reads plainly, and someone who only skims the headlines still gets the story.
4. Trust gets decided on looks before anyone reads
The people between the ATS and the design lead are looking for signals they can trust. Tim’s working definition: does it feel right, and does it make sense? A junior designer’s site used a layout, color scheme and chip style Tim recognized from a common AI site template. The designer said it wasn’t really AI-built, and Tim said that’s fine. But landing there, he’d assume the whole thing was generated, and we want your work credited to you. We almost always say the words matter more than the look. Here the look would undercut the words however good they were.
A returning designer had the opposite setup, a site full of personality aimed at a conservative industry. Nate’s read was that visual craft carries more weight when you’re targeting UI roles, so the aesthetic could be the barrier. Tim pointed to a basic UX principle: don’t surprise people. A healthcare designer’s site should look like it belongs in healthcare, and a developer-tools designer’s should look like software. SaaS teams are probably more open to a loud personality than healthcare or fintech. If your UI passes the sniff test, good writing can carry you. If it distracts, good writing won’t save it.
Keep your identity, but dress it for the room. Tim’s version was a suit and tie, or scrubs for healthcare. Save the bigger personality for the about page and the details someone finds once they’re hunting for craft, which tends to happen around the interview. Nate’s own site does this on purpose. It targets healthcare, so it’s white, blue and gray and very sterile, which isn’t his personal taste. And if someone tells you to make it look more like Apple, our advice is to thank them politely and ignore it.
5. Put the outcome at the front of every bullet
If you’re applying and hearing nothing, look at the resume before the portfolio. The ATS and the recruiter do the heaviest filtering, and the portfolio barely comes into it until later. Our last review was a resume from a designer with zero callbacks after months of applying. The layout was mostly fine. The bullets weren’t. Each one said what got done, like building a design system and rolling it out, and none said what the company got from it. As Nate put it, those bullets describe UX design. They don’t describe you.
The fix is the XYZ formula Nate shared: accomplished X, as measured by Y, by doing Z. Work out what the business got, then move it to the start of the bullet. That design system had helped the sales team sell more, so the bullet should open with that, and a conversion rate if there is one, and end with the design system. Nobody hires you for building Figma files. They’ll hire you for helping the company sell.
If you don’t have analytics to back an outcome, estimate it. Tim’s point is that the business runs on estimates too. Start with something you can reasonably guess, like how much faster the work made a team, and multiply it out to revenue, retention, churn or hours saved. Keep the inputs believable. Tim admitted his own example, ten sales per salesperson per day, was probably high. If the company is public, its disclosed numbers give you something real to anchor to.
The rest of that resume pass is worth running against your own. Put your contact details in one line under your name and title. Drop the summary. Name each role in terms of the job you want next. Cut the projects section, since it makes reviewers assume you’re a student. Remove education dates, add your location if some roles were remote, spell out months so ATS parsing and autofill work, and use a second page if you have more to say, because more words means more keyword matches. Don’t let a job description talk you down a level either. They’re often written by people who’ve never done design work. And give it time: Tim’s sense is that a typical callback takes about a month, and sometimes much longer.
