I had a portfolio review in front of me and no idea what I was actually looking at. I could tell some pieces were stronger than others. I had opinions. But when I tried to articulate why one worked and another didn’t, I kept reaching for vague language. “It feels right.” “Something’s off.” “The layout is clean.” None of that was useful. None of it could be repeated, taught, or applied to the next review.

The problem wasn’t taste. I had taste. The problem was that my evaluation criteria were implicit. They lived in my gut as feelings, and feelings don’t transfer. You can’t hand someone a feeling and say “use this to evaluate your work.”

So I started extracting.

The idea is simple. Find a practitioner whose judgment you trust, study their output and their stated principles, and reverse-engineer the evaluation criteria they’re applying. Turn those criteria into questions. Then run those questions against your own work.

Here’s the procedure. Thirty minutes, start to finish.

Step one: pick the practitioner. Five minutes. Choose someone whose work you consistently admire, and more importantly, someone whose rejections you agree with. Admiring someone’s best work is easy. The real signal is whether you agree with what they leave out, what they refuse to do, what they consider beneath the standard. That’s where the evaluation lens lives.

I’ll use Massimo Vignelli as the example because his principles are well-documented and his rejections are famously clear.

Step two: study the output. Ten minutes. Pull up five to ten pieces of their work. Don’t analyze yet. Just look. Notice what recurs. Notice the constraints they never violate. With Vignelli, you’ll see a handful of typefaces. You’ll see the grid. You’ll see color used for structure, not decoration. You’ll see a refusal to add anything that doesn’t serve the communication.

Now look at what’s absent. With Vignelli, you won’t see decorative illustration. You won’t see display typefaces chosen for personality. You won’t see complexity for its own sake. The absences define the lens as much as the presences.

Step three: extract the criteria. Ten minutes. Translate what you noticed into testable questions. This is the critical move. You’re turning observation into protocol.

From studying Vignelli, I extracted questions like: Can the number of typefaces be reduced without losing information? Is every color serving a structural purpose? Could this be simpler without becoming less clear? Is there anything here that exists for decoration rather than communication?

These aren’t Vignelli’s words. He said plenty of things about design. But the lens isn’t a list of quotes. It’s a set of questions derived from observing what a practitioner consistently does and consistently refuses to do.

Write the questions down. Five to eight is enough. More than that and the lens gets diluted.

Step four: run it against your own work. Five minutes per piece. Pick something you’ve made recently. Run each question against it. Answer honestly. Don’t defend the work. Just answer.

The first time I did this, I ran Vignelli’s extracted lens against a brand system I’d built for a client. Two of the eight questions flagged issues I’d been quietly uncomfortable about but hadn’t been able to name. A third typeface that wasn’t carrying its weight. A color that existed because the client liked it, not because the system needed it. The lens made the implicit discomfort explicit and actionable.

That’s the whole procedure. Thirty minutes. You now have a reusable evaluation instrument built from a practitioner whose judgment you trust.

But here’s where it gets interesting.

Build a second lens from a different practitioner. Someone with a different sensibility. If you started with Vignelli (restraint, reduction, structural clarity), try James Victore (personality, directness, emotional charge). The questions will be different. Is there a human being visible in this work? Does this say something that only this person could say? Is the work taking a risk, or is it playing it safe?

Now run both lenses against the same piece of work.

They will disagree. That’s the whole point. Vignelli’s lens might say: this is cluttered, reduce. Victore’s lens might say: this is timid, push harder. The same piece of work, evaluated by two legitimate frameworks, producing two different and potentially contradictory assessments.

The disagreement is where your judgment lives. You have to decide which lens applies to this work, in this context, for this audience. Vignelli’s restraint is right for a transit system. Victore’s personality is right for a poster. But a brand identity for a jewelry line? That needs both. The restraint to maintain structural coherence across touchpoints and the personality to make someone feel something when they see the mark.

The act of choosing between lenses, or blending them, or knowing which one to weight more heavily in a given situation, that’s the skill. The lenses don’t replace judgment. They make judgment more precise by giving it specific criteria to work with instead of vague feelings.

I now run six lenses regularly. Vignelli (restraint), Rand (concept as irreducible unit), Bierut (does the design think), Victore (personality and risk), Draplin (joy and craft), and a peer lens built from designers I’ve worked with directly. Each one catches things the others miss. Each one has blind spots the others cover.

You don’t need six. You need one, built this afternoon, run against something you made this week. The lens will show you what your gut already knows but can’t articulate. And once it’s articulated, you can act on it.

Start with the practitioner whose rejections you agree with. Extract the questions. Run them. See what comes back.