A brand system for a company that doesn't exist

why it exists

Two of the tools I built for my employer make the same claim: brand compliance enforced by code. That claim is invisible without a brand to be compliant with. I can't publish my employer's, so I needed one I could.

the constraint that shaped everything

It had to look nothing like my own brand. No rose, no plum, no ivory, no Space Grotesk. If a generated deck came out in my palette, the demo quietly says "my tool makes things in my style" — when the real claim is that it applies somebody else's system. The contrast between this page's chrome and a CambiumPak artifact sitting inside it is the demonstration.

Second constraint: sober mid-market B2B. Competent, unfashionable, faintly conservative. If it looks too designed, it reads as a branding exercise rather than an operations tool.

naming took longer than the mark

Ten candidates checked and rejected. Real words in an established industry are almost always taken.

The near-miss on Terrapak is the one worth remembering: same industry, one letter from a global brand. That's the failure mode — not picking a name nobody has, but picking one that reads as a knock-off of someone who has lawyers.

the mark

Four half-rings on two centres, offset along a shared vertical seam. It reads three ways at once: growth rings, a taped carton seam, and a C over a P in the counters.

The flat ends against the seam are the whole idea. The stroke caps are butt, not round. A round cap closes the gap and turns it into a generic swirl — it stops reading as a seam and stops reading as letterforms. It's one attribute and it's the difference between this mark and a thousand others.

The real constraint was reproduction: it had to survive being stamped in one colour on brown corrugated board. That kills gradients, hairlines, tight counters, and anything needing a second ink. It also has to hold at 12mm on a tape roll and 2m on a trade-show banner. Designing for the worst substrate first is why the mono lockup isn't an afterthought.

the palette, and the one colour with a job

Moss is the one to protect. It carries a single meaning — recycled-content percentages — and nothing else. A moss-coloured number that isn't a recycled percentage is a defect, not a style choice.

That rule caught something real in production. The bar charts on the spec sheets are Pine, not Moss, even though Moss would have looked better against the kraft. They aren't recycled figures, so they don't get the colour. A rule that never costs you anything isn't doing any work.

five certification marks, all invented

This is the part that mattered most to get right. FSC, BPI Compostable, How2Recycle, SFI and ISTA are real trademarks belonging to real standards bodies. Putting any of them on a fictional company's spec sheets would be trademark misuse on a public portfolio.

So they were removed and replaced with five invented ones, none confusable with a real mark.

They share a stroke weight and a corner treatment so they read as a set — which a real certification row never does, and making them cohere is what sells the whole thing.

The rule underneath them is the interesting part. A family-level badge row shows the intersection of its members' marks, never the union. If two tape codes hold ChainMark and the third doesn't, the family claims neither — it claims only what every SKU can support, and the difference is disclosed in a note. A site that shows the union is making a claim two-thirds of the range can't back. That's exactly the class of error the tooling is supposed to prevent, so getting it right here is the demo working rather than arguing against itself.

the rulebook

Rules 3, 4 and 5 are checked by the build. The site fails CI if a rendered page or the product record contains any of the five real certification body names — a check that exists because one slipped in early and I'd rather the build catch it than a hiring manager.