How we vet every designer in the network

"We only accept the top 3% of applicants" is a familiar line in this industry, and it's usually meaningless — a number with nothing behind it, repeated because it sounds impressive rather than because anyone can verify it. We'd rather describe what actually happens before someone joins the network, and let that speak for itself. Here's the real process, stage by stage.

1. Portfolio & category screen

Every applicant submits real client work in a specific discipline — brand identity, motion, web and product design, packaging, and so on — not a general "design portfolio" meant to look impressive across everything at once. This matters more than it sounds like it should. A generalist portfolio is easy to make look good by cherry-picking a handful of strong pieces across unrelated categories. A category-specific portfolio has nowhere to hide — it has to demonstrate real depth in the exact kind of work a brief would actually need.

2. Live trial project

This is the stage that actually predicts something. Not a hypothetical test, and not a take-home exercise on their own schedule — a real, timed brief with real constraints: a tight turnaround, a defined budget, a starting point that's deliberately a little ambiguous, the way real briefs often are. It's evaluated blind by a senior art director who doesn't know who submitted it, which matters as much as the test itself — a name or a reputation can quietly bias a review, and this stage is specifically designed to strip that out.

Portfolios show what someone is capable of when they control every variable. A live trial shows what they do under the actual conditions client work happens in.

3. Revision & direction test

This one surprises people, because it's not testing design skill at all — it's testing something else entirely: how someone responds to feedback. A separate short exercise gives a designer a note or a piece of art direction and evaluates how they interpret it, whether they revise in the right direction, and whether they lose the thread of the original brief in the process.

Here's why this stage exists: raw talent and being genuinely good to work with are two different things, and portfolio review only ever measures the first one. Someone can produce beautiful work and still be difficult to direct, defensive about feedback, or prone to "fixing" the wrong thing when a client asks for a change. This stage is where that shows up — before a real client ever has to find out the hard way.

4. Reference check

Verification of real history with past clients or agencies — not just confirming the work in the portfolio is genuinely theirs, but asking a more useful question: did the people who worked with them want to keep working with them. A designer can have an excellent portfolio and a track record of clients who quietly moved on after one project. References are where that pattern becomes visible.

Portfolio review asks "can they do good work." The trial and revision stages ask "will they do good work here, under real conditions, when asked to change direction."

5. An ongoing standard, not a one-time gate

This is the stage most vetting processes skip entirely, and it's arguably the most important one. A lot of vetting elsewhere is treated as a single hurdle — clear it once, and you're in permanently, regardless of what happens afterward. That's not how this works.

Every single delivery that goes out — regardless of how long someone has been in the network, regardless of past track record — is re-checked by both a human art director and AI before it ships. Inclusion in the network isn't a status that, once granted, is permanent by default. It's maintained by consistent output, project after project, the same standard applied evenly whether it's someone's first week or their fiftieth project.

Why we don't cite a number

We could tell you an acceptance percentage. It would be easy to write and impossible for you to verify, which is exactly why we're not doing it. What we can actually stand behind is the process itself — five real stages, each testing something specific, described honestly rather than compressed into a marketing statistic. If that process changes or gets more rigorous over time, this page will be updated to reflect it, not left stale with a number nobody's checking.

← Back to Journal