I’m Diana Lopez, a product engineer and the founder of PixelsWithin, a software studio where I design and build websites and apps for businesses. Most of the people who end up working with me don’t come to me with a spec. They come with a feeling.
If you’ve been told by an agency or a developer that they need a full brief, wireframes, or a requirements doc before they can start—and you don’t have one, because you don’t fully know what you want yet, only what it needs to feel like—this is for you. “I want the site to feel more premium.” “I don’t know, something more us.” “I’ll know it when I see it.” The details aren’t the thing that’s missing when a client talks like that. The words are.
This particular example is a design one—most of what follows is about brand and tone, not code. But I want to be upfront that I don’t usually experience my work in separate lanes like that. Most clients come to me with the whole kit and caboodle: the brand, the product, the build, all at once, all tangled together. I’m walking through the design side of it here because it’s the clearest version of this particular story, not because it’s the only place this kind of thinking shows up.
Here’s a technique I’ve used for years to get from a feeling to a decision, and the strange, small business it accidentally turned into.
The exercise
When a client can’t articulate what they want their brand or product to feel like, I don’t ask them to describe it. I hand them a block of adjectives—a big, messy list of tones and personalities a brand could have—and I ask them to do two things, in order.
First, cross out everything that doesn’t resonate at all. Not “pick your favorites”—the opposite. Get rid of what’s clearly wrong first, because most people find it much easier to know what they aren’t than to know what they are.
Then, once the list is smaller, circle the ones that are left that actually pull at something. Not the ones that sound nice. The ones that feel true.
What’s left after those two passes is usually a small, specific handful of words—and almost every time, the client looks at that shortlist and says some version of “yeah, that’s it.” They didn’t have the words a minute ago. They do now. And once you have words, you have a goalpost. Every decision after that—a color, a layout, a line of copy, a tone in an email—can get checked against them. Does this still feel like the words we circled? A feeling can’t be checked against. A word can.
I’ve run some version of this exercise with dozens of clients over the years—you can see some of that work here. It’s rarely the same twice, because the list changes and the client changes, but the two passes—cross out what’s wrong, circle what’s true—hold up almost every time.
This is, I think, part of why clearly-worded input matters even more now that AI tools are part of how a lot of creative and technical work gets done. A model can do a lot with “editorial, a little irreverent, quietly expensive.” It can’t do much with a vibe you haven’t said out loud yet.
How I know it actually works
I didn’t just use this exercise a few times and assume it was good. I know it works because it turned into a business.
I originally wrote it up as a free PDF—just the adjective list and the two-pass instructions. It sat online for years before it started gaining real traction in search. Once it did, I noticed something else happening alongside the traffic: people were stealing it. Someone lifted my list wholesale and reposted it as their own.
Instead of being annoyed and leaving it there, I thought: fine, I can do this better anyway. I turned the static PDF into an interactive quiz and built it out into its own site, How To Brand You. I redirected the original article straight into the new site, which meant the SEO equity the PDF had already earned carried over instead of starting from zero.
About 1,400 people ended up taking that quiz before I eventually sold the site—to a copywriter named Andrew McKay, who did a lot of work with brand archetypes, which was one of the frameworks I’d built into the tool. The traction wasn’t a guess or a hope. I could see it in the numbers, and eventually in the fact that someone else wanted to own it.
What this means if you’re the one without the brief
I’m not telling you this story because everyone who hires PixelsWithin needs a brand-adjective quiz. I’m telling you because it’s a concrete example of the actual mechanism I use whenever someone comes to me with something they can’t fully explain yet: get rid of what’s wrong first, find out what’s left, turn a feeling into a word, and use that word as the goalpost everything else gets measured against.
That’s usually where the real work starts—not after you’ve handed me a perfect brief, but before you have one.