Every route is now cut in a condensed grotesque, and every route got pushed past the one-shot idea into something with a second and third move. The wide Helvetica version is still in the menu as a control.
ASSEMBLISTS is eleven letters — long enough that a wide face forces the mark into a thin horizontal band that dies in every square format. Condensing buys vertical mass: more black, tighter counters, a denser object. Brutalism is about density, and density is what a narrow cut gives you for free.
The buttons in the top bar recut every mark on the page live. Layout is measured at render time, so each route re-solves its own proportions for the face you pick — nothing is hard-coded to Helvetica.
Start at the type lab below to pick a face, then work down. Each route has a black Developed band under the main mark holding the extensions. ↻ Replay re-runs any animation.
All seven are set to an identical cap height, so the only variable on screen is width. The number on the right is the wordmark's width as a percentage of the widest cut — a real measurement taken at render time, not an estimate.
HN Condensed Black for the marks that need mass (02, 04, 06) and DIN Condensed for the ones that trade on precision (01, 05). DIN is the German industrial standard drawn for machine parts and signage — for a route about milled slots and tenon joints it isn't a style choice, it's the correct dialect.
Very narrow faces bleed legibility fast at small sizes and in reverse. Impact and DIN both close up when knocked out of black. Check every route's Reversed strip before you fall in love with a width.
These are all system faces standing in for the real thing. Production equivalents in the footer. Route 03 is exempt — it has no typeface at all.
The watch-out from v1, answered. Below ~30px the overshoot and the channel walls are dropped and the rod sits flush to cap height. Same idea, one third the parts. Shown at 100 / 48 / 24px.
The exclusion zone is derived from the mark rather than invented: one rod-width on every side. It scales automatically with the logo and it's a rule anyone can measure without a spec sheet.
Once the rod exists it stops being decoration: it's the bullet, the section divider, and the progress marker. The identity is one component used at four scales.
ASSEMBL·I·STS. The I is the only letter in the word that is already a pure rectangle — a bar, a part, a component. So it's removed from the typography and put back as an object: a machined rod dropped into a milled channel.
Everything improved. In a condensed cut the letters are closer to the rod's own width, so the rod stops reading as an intruder and starts reading as one more vertical in a rank of verticals — present, but native. In DIN it's almost invisible until you look, which is the right amount of loud.
The act of assembly happens inside the name, using a piece the name already contained. Nothing is added — a letter is re-manufactured. That's the whole consultancy pitch in one glyph: your parts, engineered properly.
The rod is sized off the face's own stem width, so it survives a typeface change — but it does not survive a weight change. If marketing ever sets the wordmark in a lighter weight the rod will look bolted on. Lock the weight in the guidelines.
The development a narrow face unlocks: ASSE / MBLI / STS lands at roughly 1:1.15 — near enough square for an avatar or an app icon. In wide Helvetica the same three lines produce a block half again as tall.
One more break gives a 1:2 portrait for banners, spines and vertical formats. It won't go square — eleven letters over four lines can't — and legibility is genuinely marginal here. Included so you can see where the idea stops.
The real extension. Force-justification isn't a logo trick, it's a layout rule — any message set the same way inherits the identity. The logo stops being a stamp and becomes the house style.
The word is broken at the wrong place and force-justified into a solid black slab. You reassemble it in your head. That's the joke and the job.
Lines are set to identical width by changing point size, never by stretching glyphs — so the cap heights differ and the block stays a perfect rectangle. The white kerf runs edge to edge: a saw cut, not a gap.
Changed what the route is. Wide, it was a two-line horizontal bar. Narrow, the line count becomes a variable and the same construction yields a bar, a near-square and a portrait — one idea covering nearly every aspect ratio you'll be asked for.
Justified stacks live or die on the break points. ASSE/MBLI/STS is deliberately arbitrary, and some people will read that as a mistake rather than a decision. It needs confident deployment or not at all.
The point of drawing your own letters: narrow is a number you type, not a font you licence. Same glyph logic, same stroke, three grid widths. Pick a width per application instead of buying a family.
The entire alphabet is one bar in four lengths. Every glyph in the name is between one and six placements on the grid — that's the whole specification, and it's why the width can be a variable.
0–9 need no new modules, which is the test of whether a system is real. A–Z is next; K X Y Z will need a wider grid or a diagonal module, and that's the honest open question.
Don't set the name — build it. A purpose-drawn stencil alphabet on a strict grid: one module, no curves, no diagonals, monospaced advance with a single half-width exception for I.
It became the argument for the route. Condensing a licensed face means paying for another cut and hoping it exists; condensing this one means changing a single number. Route 03 is the only proposal here that treats width as a parameter you own.
The only route producing proprietary letterforms rather than an arrangement of someone else's. It says "custom tools from common parts" by being exactly that, and it scales into a headline face, an icon set and a pattern.
Rectilinear stencil carries a whiff of 1970s wayfinding — an asset if that's the intended read, dated if not. And it is a genuine typeface project: budget drawing time, not just a logo fee.
The tenon is defined off the mark's own height — depth ¼H, width ⅓H, centred — so it survives any recut. Every brand element cut to this profile interlocks with every other one. That's the deliverable, not the logo.
Rotate the profile and it stacks: service cards, slide sections, pricing tiers. A narrow wordmark is what makes the vertical case viable — the wide cut produced modules too letterboxed to stack.
Three notched modules carrying real content. The logo is now the smallest instance of the layout grid rather than a thing placed on top of it.
The logo isn't a picture, it's a connector. A black slab with a tenon on the right edge and a matching mortise on the left, so it locks into copies of itself — or any other element cut to the same profile.
Made the module usable. Wide, the slab was 6:1 and only ever tiled in a horizontal strip. Condensed brings it near 4:1, so the same joint now works stacked, gridded and as a card — the system got a second axis.
Composability made physical. A consultancy that assembles tools from interoperable parts should own a mark that is itself an interoperable part — and it leaves designers a rule to follow after you've gone.
The asymmetric silhouette fights centred layouts, and embroidery and etching will want a de-notched version. Specify the plain rectangle as an official alternate rather than letting people improvise one.
The same construction at five offsets. Zero is a solid wordmark; increasing the offset opens the letterforms into structure and eventually into noise. Pick a fixed value for the logo and use the rest as a motion and layout range.
This route's flaw was fragility. A condensed cut has thicker stems relative to its counters, so the surviving slivers are chunkier and the mark holds together at offsets — and sizes — where the wide version fell apart. Top: wide. Bottom: condensed. Same offset.
Offset zero is the fallback, which is the elegant part: the small-size, embroidery and single-colour version isn't a different drawing, it's the same mark with the variable set to nought.
One solid word, printed twice, one impression off-register. Where the two hit they cancel to white; where they miss, black survives. A solid block becomes a hollow structure — the outline is the residue of the offset.
Not an outline font and not a drop shadow: genuine difference blending. The white ghost knocks out to white over black and prints black over the white ground, which is a true XOR. Drag the slider above to drive it.
Fixed the weakness. Thicker relative stems mean the residue is a solid structure rather than a hairline, so the mark survives further down in size. It's still the most fragile of the six, but it's no longer disqualified.
It reads as a treatment more than a mark. Best used as the expressive register — campaigns, covers, motion — over a plain wordmark doing the everyday work. Don't put it on the invoice.
Five courses instead of seven, so each step is bigger and the kerfs survive further down. At 16px they fuse into a solid A — the correct failure, degrading into a legible letter rather than mush. No separate small-size drawing needed.
One rule governs both: the mark is a square whose height equals the wordmark's cap height, baseline-aligned. Recut the type and the mark re-sizes itself — no eye-judging, no second spec.
The same five-step rule builds rules, dividers and data bars. Fewer, heavier courses make the device read at a glance where seven thin ones turned into texture.
Everything else here is a wordmark, which leaves nothing for an avatar, an app icon or a favicon. This route answers that: an A built from five courses stepping out toward the base — a structure under construction that happens to spell the first letter.
Outer width steps out a constant amount per course; the counter steps identically, offset by a constant stroke, so both edges diverge in lockstep and the letter keeps even weight. Course four of five is left solid — that's the crossbar. Nothing is drawn but rectangles; every counter is a cut.
The symbol is drawn, so the menu doesn't touch its shape — but it now takes its size from the type. Tying mark height to cap height means every recut re-proportions the lockup automatically, and a condensed cut keeps the square mark from looking stranded beside a long word.
Abstract A-marks are crowded territory; this one earns distinction from the leg gap and the missing apex, so those proportions are load-bearing. Deploy it with a wordmark, not instead of one.