JAFFX

The name was already there: Joel A. Jaffe, plus FX. Everything else about how the company looked and read was open. No designer had ever worked here, so there was no house style to inherit and nothing to argue with.

There was also nobody to check my work. The founders trusted my judgment completely and never pushed back on a single decision in eleven months. I expected that to feel like freedom. Mostly it felt like weight: every consequence of every choice was going to be mine, and if something was wrong, no one was going to catch it. That shaped how I worked, and it comes back at the end of this study.

• Pedal Curious

  Little or no experience with effects. Interested, but analog pedals mean buying five units to get a usable range of tones, and digital multi-effects units bury everything in menus.

• Tinkerers

  Care about customising their sound as much as playing it. Their problem is the opposite one: existing products are not extensible, and nothing talks to anything else.

The product answered this with one device that does both, analog-feeling to play and reprogrammable underneath. The brand had the same problem and no hardware to hide behind. Whatever I made had to be legible to both people at once, in a single glance, on a poster.

• Boutique analog

  Hand-painted enclosures, illustration, nostalgia. Wins the Pedal Curious and reads as a novelty object rather than a programmable platform.

• Digital multi-FX

  Black plastic, grey LCD, dense and severe. Wins Tinkerers by looking like the menu-diving problem we existed to solve.

• Open-source tools

  Monospace, terminal, flat and utilitarian. Credible to Tinkerers and invisible to a beginner. No warmth, and no music in it.

Each of the three adjacent worlds solves for exactly one of my two customers. Borrowing any one of their languages would have sold the product to half its market and quietly written off the other half. The brand had to sit in the gap none of them occupied, which is why sleek, fun and warm was not a mood board. It was the position, and it was the only one that reached both segments.

• Sleek

  Simple geometric shapes, gradients, and #1A0D1C as the standing ground. Monospace for everything you read. This is what buys credibility with people who write code.

• Fun

  Variety of shape, bold colour, texture. The pedal should be as fun to explore as a toy and powerful enough to say something with. This is what lowers the cost of picking it up.

• Warm

  One colour carries it. #F14E13 is the primary, and it is doing the work of making an open-source project feel like a place rather than a repo.

The monospace decision is the one I would defend hardest. Mono is the native typography of code and open source, so every line of text on the site quietly tells a Tinkerer that you can get inside this, while staying calm and readable for someone who cannot. Expression is concentrated in the display letterforms and the gradient, and the layer you actually read stays technical. One rule, and most of the tension resolves itself.

The second rule lives in the navigation. Nav items are pedal knobs and the active state is the status LED, so the playfulness sits in the interaction metaphor rather than in decoration. Fun that costs nothing in sleekness.

A free face with unclear terms was a risk I did not want to hand the founders. Drawing our own removed the problem and gave the company something it owned outright. The first directions were wrong in an instructive way.

• A rounded, bulbous rock wordmark. Warm and immediate, and completely wrong: it reads as a band's logo, sells nostalgia to the Pedal Curious, and says nothing at all to a Tinkerer.
• Literal music iconography, a treble clef and staff notation. Wrong instrument for a guitar product, and notation signals learn to read music first, which is the exact intimidation the product exists to remove.
• JFX, as a shorter mark. Cleaner, but it threw away the founder's name, which is where the whole thing came from.

What survived was built from the sleek rule and the fun rule at the same time: letters cut from solid geometric slabs, each broken by a diagonal negative-space gap. The geometry is the sleek half. The cuts are the fun half.

Wide, uppercase and stencil-like, it borrows from amp faceplates and roadcase stencils rather than from posters. Two honest notes about how it was made. I drew the letterforms as vectors in Figma, in the same file as the rest of the system, rather than cutting a font file. That kept the whole brand in one place while it was changing weekly, and it is why the hardest part of the work was spacing: with no kerning table, certain pairs had to be corrected by eye on every application, and a few never fully settled. Cutting a real font with proper kerning pairs is the first thing I would do next.

#F14E13 is the primary and the carrier of the warmth. Between them runs the gradient into periwinkle that appears in the wordmark and across every surface. Everything else is an accent, held to a supporting role. Accents earn their place when they carry information rather than decoration.

• In the pitch deck

  Each part of the value proposition takes its own accent: mint for user-friendly, gold for infinitely extensible, periwinkle for open-source. The palette is encoding the three things the company claims.

• In JAFFX Studio

  Every effect block is a distinct gradient tile, so a saved signal chain is readable at a glance without reading a single label. Colour is doing product work.

No budget, no other designer, no existing assets. The slide I would point at first is the market-size chart. TAM, SAM and SOM is the most standardised and least interesting diagram in venture pitching, and I drew it as nested organic waves in the brand palette with a plectrum-shaped marker for the beachhead. It is the fun rule surviving contact with the least fun content in the company, which is the only way to find out whether a principle is real or decorative.

The system's real test came later and without me. Our CEO used it to assemble grant applications. Marketing used it to make videos, flyers and posters.

I was the frontend lead as well as the designer, and shipped jaffx.audio: home, demo, team, studio and contact, with the waitlist and the icon set I had drawn. The team's own market validation had concluded that the website was a component of the product rather than marketing around it, which is why it mattered that one person did both. Nothing was lost in handoff because there was no handoff. The tokens in the design file and the custom properties in the stylesheet were the same decisions, written twice.

44

Teams at the start, October 2024

6 of 20

Through the New Venture Fair, judged in trade-show format on booth, demo and poster

2nd

At the finals, in front of independent judges. $7,500, on a $250,000 prototype-phase ask

The middle round is the one that mattered for this work. The Fair is judged on a booth, a demo and a poster, so at that stage the visible product was largely the brand. Twenty teams went in and six came out. I will not claim the identity caused that. I will say that the round where my work was the thing being looked at is the round we survived.

Nobody pushed back on anything, so nothing caught what drifted, and three things did. The palette grew past what one person could reasonably govern, and ten swatches with no stated hierarchy is a system that will be misread the moment someone else opens it. The principles lived in my head rather than in a document, which is why the brandbook records a type stack and a colour grid instead of the attribute-to-mechanism map I was actually working from. And the letterforms stayed Figma vectors, so spacing needed correcting by hand on every application. All three are the same failure in different clothes.

I would write the principles in week two, when they are cheap, instead of reconstructing them afterwards. The one move JAFFX never got to is the one I would want most: it is a music company with no sound and no motion in its identity, a brand for people who make noise, that has never made any.