We killed our own parent brand before it shipped
Code One was supposed to live under Atrivo. Before a single production board shipped, we dropped the parent brand entirely — and it was a QR sticker that forced the call.
This week I filed a trademark application for Code One. In the same breath, I killed Atrivo — the parent brand Code One was supposed to live under. Nothing with Atrivo on it had shipped at scale yet.
That was exactly the point.
One product doesn't need a parent
Brand architecture — a house name sheltering product names — is a tool for companies with portfolios. P&G needs it. A company with one keyboard does not.
Every touchpoint that said "Atrivo" was asking customers to remember a name that bought them nothing. Nobody searches "Atrivo." They search code one keyboard. They recommend "the Code One." The parent name was a third brand token soaking up the very limited attention a niche brand gets, and the customer's mental model had to hold a relationship — Atrivo makes the Code One — instead of a single fact: Code One is a keyboard for people who code.
Small brands win by being one crisp idea. Two names is two ideas.
The name was already doing the work
"Code One" isn't a neutral label. It says who it's for. It carries a flagship connotation — the first, the primary, the one. The keyboard's design language leans into it: the accent colour lives on the #1 key. Name and product reinforce each other without a parent brand in the middle explaining the relationship.
Even the domain got sharper. getcodeone.com is a call to action. atrivocode.com was a compromise.
The tax you don't see
The cognitive load argument usually gets made about customers. But we were paying it internally too. Every artwork file, email signature, social handle, and piece of copy had to answer the same question: which name leads here, and how do they relate? A small tax, paid on every single asset, forever.
One brand means the question never comes up again.
A sticker forced the decision
Here's the part I find funny: the thing that finally forced the call was a QR code sticker.
The back of every Code One has a blanked-off port — a knockout with a faint witness line where an opening would have been. Left bare, it reads as an unfinished generic case. So we're covering it with a matte black QR sticker, serialized per board, and that meant answering a question with real permanence: which URL do you tattoo on a thousand keyboards?
A URL implies a domain. A domain implies a brand. And suddenly a vague "we'll probably simplify the branding eventually" had a hard deadline: the artwork lock before production. So the call got made, the trademark got filed, and the sticker artwork will carry the answer.
Decisions like this have closing windows. Ours was the final pre-production artwork round — after that, the wrong name gets printed a thousand times and the cheap decision becomes an expensive one. If you have a fuzzy strategic call you keep deferring, look for the physical artifact that will freeze it. That's your real deadline.
What it costs, honestly
Killing a brand pre-launch is cheap, not free:
- Every factory-bound file needs a sweep. Box artwork, unboxing card, manual, PCB print, sticker template. One missed file and the old name is baked into the entire first batch.
- The old domain lives on as a redirect for years. Dropping a brand from your identity is not the same as abandoning its domain. A lapsed domain someone else picks up is a dumb, avoidable risk.
- Common words face more trademark scrutiny. "Atrivo" is invented — invented names sail through examination. "Code One" is two common words, and filing is not the same as granted. We hold both marks through the transition.
- Email migration has a tail. The support address is wired into store notifications, tooling, and customers' address books. That's a planned migration, not a side effect.
The nuance we're keeping
Pure Code One means keyboard-centric, not keyboard-only. Companion accessories, the tools we want to build around typing — all of it still fits, because all of it orbits the keyboard. What died is the holding-company layer, not the ambition.
And yes — you're reading this on atrivolabs.com. The lab keeps the name. The product never says it again.