Your portfolio gets read second

Designers talk about portfolios. We compare case study structures, argue about how much process to show, redesign our own sites every eighteen months.

Meanwhile, in most application flows, the first file that actually gets opened is the CV.

It is the shortest document you own and the one that gets the least attention from you. And it is often telling a different story about you than your portfolio does.

I wanted to look at that gap in a concrete way, so I used the free CV scorer we built at Folioverse. Below is what it says about an example CV, dimension by dimension.

A note before we start, because it matters for how you read this. The CV belongs to Maya Reyes, a fictional designer whose portfolio we published as a worked example. We wrote her CV ourselves. The assessments quoted below are not written by me either; they are the tool’s actual output on that text. So this is not evidence that anything works. It is an illustration of what a structured read of a CV looks like.

The overall result was 58 out of 100, with this summary:

Clears the style bar but stalls on outcomes: no metrics anywhere means it sits in the ‘maybe’ pile.

Here is how that breaks down.

1. Impact and outcomes

The question it asks: did anything change because of your work, and can you say by how much?

This is where most design CVs collapse, including this one. It scored 38, the lowest of the five.

Every bullet describes what was built, not what changed because of it, no metrics, no before/after numbers anywhere.

Look at a line from Maya’s CV: “Redesigned the mobile Send Money flow for clearer confirmation and fewer costly errors at the moment of transfer.”

That sentence is well written. It names the surface, the intent, and the stakes. It still does not say whether anything improved.

The distinction is not about having impressive numbers. It is about whether the reader can tell the difference between a designer who shipped a change and a designer who shipped a change that worked. Without a result, both look the same on paper.

If you have no clean metric, a qualitative one is still a result. Task completion in a usability round. Support tickets on that flow. Time to complete. Adoption in the first month. Even “five out of five participants completed it on the first attempt” tells the reader more than an adjective does. The same logic carries over to case studies, and I wrote about it in how to write a UX case study when you don’t have metrics.

2. Role and scope clarity

The question it asks: what was yours, and how big was the room?

Scored 68.

Senior vs. mid level is implied by title progression, but team size, stakeholder count, and individual ownership are never stated.

This is the “I versus we” problem. “Led the redesign” reads differently if you were the only designer on a four person squad than if you were one of five designers on a platform team. Both are legitimate. Only one of them is what the reader will assume, and you do not get to choose which. It is the same ownership question that comes up in showing your contribution in a team project, compressed into one line.

One parenthetical per role usually fixes it. Team size, whether you were the only designer, who you worked with.

3. Structure and scannability

The question it asks: can a reader, or a parsing system, find the important things quickly?

Scored 72.

Clean structure, consistent dates, and a portfolio link, scans well, though the skills section is thin and may miss keyword filters.

I want to be careful here, because this is the dimension people turn into a game. There is a whole category of tools built around gaming applicant tracking systems, and that is not what this is about. Consistent dates, a clear role line, a portfolio link that works, and a skills section that actually names what you do are not tricks. They are the difference between a document that can be read quickly and one that cannot.

4. Craft and specificity

The question it asks: does the CV show how you work, or only that you work?

Scored 65.

The writing signals taste and systems thinking, but the tools list is skeletal and no research methods or deliverable types are named.

A tools list of four items tells the reader nothing except that you have a Figma license. What the reader is looking for is method. User interviews, usability testing, journey mapping, design critique facilitation. The named methods do more for you than the named tools, and they take one line.

5. Narrative and positioning

The question it asks: after reading the top third, does the reader know who you are?

Scored 74, the highest of the five.

The summary lands a clear identity (systems thinker, fintech/B2B, 0-to-1 capable) and the role arc reads as deliberate progression.

This is the part most people overwrite and then delete. A summary that names a domain, a way of working, and a scope does real work. A summary made of adjectives does not.

The read behind the score

The tool also writes a simulated reviewer read, which I find more useful than the numbers. This was the “what creates doubt” section for Maya:

I finish reading and I have no idea if any of this worked. Did the Send Money redesign reduce transfer errors? By how much? Did the dashboard redesign change how operators use the product? Without a single number, I can’t distinguish this from a designer who shipped pretty screens that nobody used.

That last sentence is the whole problem in one line. The CV is not failing because it is badly written. It is failing because it is indistinguishable.

What the fix looks like

The tool’s first suggested fix was to add a measurable outcome to every role, and it offered this as illustrative phrasing:

Redesigned the Send Money confirmation flow, reducing user-reported transfer errors by ~30% in the first month post-launch.

Same work, same sentence structure, one clause added at the end. That clause is the part the reader was waiting for.

Which is worth noticing: the work was already done. The result existed. It just never made it into the document. This is the pattern I keep seeing in portfolios too, and it is the same pattern. The invisible half of the work stays invisible because nobody wrote it down.

Where a score helps and where it does not

A score is not a verdict on you. It is a first read, done consistently, on a document you have looked at so many times that you cannot see it anymore. That is its only real advantage over your own judgment: distance.

What it will not do is tell you what to say. The outcomes have to be yours, remembered and dug up from old decks and old Slack threads. No tool can invent them, and any tool that offers to is inventing them.

If you want to see the five dimensions run on your own CV, the scorer is free and does not require an account. Paste the text or upload the PDF.

And once the CV is doing its job, the harder question comes next, because a strong CV gets you the interview and the portfolio behind it is what gets you the offer. That is a different piece of writing, and I will come back to it.

FAQ

What are the five dimensions a CV can be read in?

Impact and outcomes, role and scope clarity, structure and scannability, craft and specificity, and narrative and positioning. Each one is a question a reviewer is silently asking of the document.

What if I have no metrics to put on my CV?

A qualitative result is still a result: task completion in a usability round, support ticket volume on a flow, adoption in the first month. The goal is to show that something changed, not to show a large number.

Is this about beating ATS filters?

No. Consistent dates, a clear role line, and a skills section that names what you do make a document readable for humans and parsers alike. Tricks aimed at gaming tracking systems are a different activity, and not one this article recommends.

Is the CV in this article real?

No. Maya Reyes is a fictional designer we published as a worked example, and we wrote her CV ourselves. The quoted assessments are the tool’s real output on that text.

Does a low CV score mean my application is doomed?

A score is a consistent first read, not a verdict. Its value is distance from a document you can no longer see clearly. What you do with the gaps it points at matters more than the number.