Most advice about logo lettering starts in the wrong place. Designers fall in love with a sketch, polish it until it looks good on a mood board, and only then discover the wordmark breaks the minute it has to live on a mobile header, a carton panel, a broadcast ident, or an embroidered patch.
Lettering for logos isn't a drawing exercise first, it's a system-design problem first.
That's why I treat hand-drawn letters as engineering disguised as drawing. The shape matters, but so do the tolerances, the spacing, the scale behaviour, the fallback files, and the rules that stop the mark collapsing when other people start using it.
UK identity history shows why this matters, from the Transport typeface on road signs to the British Rail double-arrow introduced in 1965, where lettering had to work in public, at speed, and under severe legibility constraints, not just in a presentation deck (mid-century British typefaces and fonts).
Table of Contents
- Why Most Logo Lettering Projects Fall Apart
- Sketching Ideas on the iPad
- Moving Sketches Into Adobe Illustrator
- Spacing, Kerning, and Optical Rhythm
- Using Client Feedback to Sharpen the Style
- Locking the Wordmark With Vectors and Glyphs
- A Lettering Brief Checklist for Brands
Why Most Logo Lettering Projects Fall Apart
The project usually fails after the sketch looks “right”. A client approves a lively rough, then production exposes the gaps.
Stems misalign when the wordmark shrinks, counters clog, alternates were never drawn, and the logo exists only as one lonely PNG that can't adapt to anything beyond a single use.
The real failure point is the system
Once a mark leaves the studio, it has to survive websites, signage, packaging, motion, and print. UK brand guidance makes that operational reality explicit with minimum logo widths of 30 mm in print and 70 px on screen, plus exclusion zones at least the height of the uppercase letters, alongside warnings against busy backgrounds, stretching, and altered proportions (UK brand guidelines). If you never planned for those conditions, the logo is already half-broken before launch.
Practical rule: if the letterforms only look good at one presentation size, they're not finished.
The deeper issue is that custom lettering carries typographic decisions that other teams will inherit. Weight range, cap height, x-height, sidebearings, and the way the counterspace behaves all need to be decided early. That's the difference between a mark and a usable identity asset.
British lettering history backs that up. The BBC's 2021 move from Gill Sans to its own custom BBC Reith Sans, and Fortnum & Mason's 2025 reveal of a fully bespoke typeface family after 300 years, both point to the same business logic, that major institutions use custom letterforms to control voice, heritage, and clarity across channels (history of typography and bespoke branding). When the system is designed properly, the logo survives contact with the world.
Sketching Ideas on the iPad
I start on an iPad because it keeps the early phase loose. The point is to make enough options to see the shape of the problem, not to perfect a single answer too early. If you're using an app setup that supports fast sketching, layering, and quick export, this roundup of top iPad apps for artists is a useful reference point.
Make procreate do the heavy lifting
I sketch in Procreate on an iPad Pro with a textured pencil brush. I'm looking for forty to sixty thumbnail explorations, spread across at least four directions, before I judge anything properly.
Those directions need to feel different. I'll vary the construction grid, the stress, the terminal shape, and the contrast on purpose so I'm not just drawing the same idea with minor changes. Every page gets short notes, the intended word, the grid logic, and any instinct about how it should feel later in vector form.
Keep the first pass rough enough that you're still making decisions, not just colouring in a favourite.
I also avoid redrawing favourites cleanly at this stage. That habit makes weak ideas look stronger than they are and wastes time on a direction that may not survive production anyway. The job here is to search, not to finalise.
Export only the strongest pages
When the pages are done, I export the three strongest as layered PSDs and PNGs. That gives me enough structure to bring them into Illustrator without locking myself into one route too early. If a sketch can't hold its shape when handed off, it wasn't ready for the next stage.
A useful side effect of this process is that the file stays searchable later. Months down the line, when a client asks why a certain terminal was chosen, the notes are still there.
Moving Sketches Into Adobe Illustrator
The mark stops being an idea and becomes a working vector. Adobe UK's lettering guidance is clear on the sequence, draw by hand first, then digitise in Illustrator, vectorise, refine readability and kerning, and export the result for reuse (Adobe Illustrator lettering guidance). I follow that path closely because it keeps the sketch honest.
Build three clean concept directions
Each sketch file goes into its own artboard at 1:1 scale. I lock the scan or PNG as a template layer at roughly 50 percent opacity, then trace the structure with the Pen Tool. I use the original drawing as a reference, not a crutch, because hand-drawn looseness often hides inconsistent stems and uneven joins.
For the counters, I'll combine Pathfinder and Shape Builder rather than relying on automatic trace tools. Image Trace is quick, but for production lettering it tends to chase pixels instead of construction lines. That usually flattens the character you wanted in the first place.
After that, I redraw every letter against a baseline grid with consistent overshoot and stem weight. The result isn't a final brand yet, it's three clean black-and-white vector wordmarks ready for spacing work. I save separate AI files named by direction, then export each as SVG so the client can review them without getting distracted by styling.
Why the output stays monochrome at this point
Colour too early hides problems. A wordmark that depends on tint or texture to feel balanced hasn't solved the underlying shape. I want to see whether the skeleton works before anything decorative gets a vote.
The practical toolset here is simple, Adobe Illustrator for the vector pass, and Glyphs later when I need to test the wider type system. That combination keeps the production chain tidy. It also makes it easier to hand off files to another designer without losing the logic.
Spacing, Kerning, and Optical Rhythm
Spacing is engineering, not taste. A clean outline can still read badly if the rhythm between letters is off, and that problem gets worse once the mark is compressed into a small header or pushed onto a busy package panel. Research from the University of Reading shows that recognition changes with context, and that simpler designs, open counters, wider letter shapes, and wider spacing improve legibility, especially for low-vision audiences (University of Reading research review).
Start with the metric discipline
I build a metric grid in Illustrator and set units to thousandths of an em. Then I define a sidebearing that matches the stem weight and drop in a system sans, often Arial, as a placeholder to check balance before custom kerning starts. That gives me a neutral frame of reference before I begin making the letters behave like one word.
Kerning happens in three passes. First, I let the metrics establish a mechanical baseline. Then I fix awkward pairs like AV, VA, To, Ty, and LT. After that comes the optical pass, where I correct the illusion-driven gaps that make round letters sit too close to verticals.
| Element | Target tolerance | Check method |
|---|---|---|
| Pair spacing | No pair more than a small deviation from the average tracking | Compare against the spacing master and scan the word as a whole |
| Word spacing | Keep the wordmark compact without crowding the counters | Test at the intended smallest size |
| Silhouette rhythm | The word should read as one unit | Squint test, then grey the type to 40% |
| Small forms | Narrow letters like h and i must stay distinct | View at distance and speed, then recheck at scale |
If the silhouette breaks when you grey it down, the spacing isn't solved.
Lock the spacing before polishing the rest
I don't touch colour or weight decisions until the spacing master is fixed. That sounds strict, but it saves a lot of rework. Once the rhythm is stable, the rest of the system can be tuned with confidence instead of guesswork.
Using Client Feedback to Sharpen the Style
Client feedback usually gets treated like interference. It's better to treat it as a diagnostic. When someone says the lettering feels too friendly or too sharp, they're pointing at a variable, even if they don't name it that way.
Translate taste into typographic levers
I map comments back to stroke contrast, terminal shape, x-height, or aperture. If a stakeholder says, “I just don't like the A,” I push for a comparison. Which A should it resemble instead? That forces specificity and stops the conversation from becoming a battle of instincts.
The same rule works when the brief gets fuzzy midstream. A client may notice something I missed because they live with the brand every day. That's useful information, not a nuisance, as long as it gets recorded in the right language.
Present decisions, not just revisions
I show revisions as side-by-side pairs. That way the client can see the trade-off being made instead of feeling like they're being asked to bless a mysterious change. After two rounds, I write a short style note with counter size, contrast, and the agreed personality keywords, then I keep that document next to the vector files.
If the brief needs a larger type system rather than a one-off mark, I'll point teams to a custom font workflow, including how to commission a custom typeface, so the lettering and the wider brand voice stay aligned.
This is also where the work starts to feel more like brand governance than illustration. The client is no longer judging a drawing, they're approving a system that will shape how the brand sounds in every place the wordmark appears.
Locking the Wordmark With Vectors and Glyphs
The approved sketch has to survive a 12 mm favicon and a 12-metre hoarding without losing character. That's why I trace the final version in Illustrator with Bézier handles kept as consistent as possible, then move the masters into Glyphs 3 to test spacing, hinting, and weight consistency. Adobe's own UK guidance also makes the technical point plainly, custom lettering becomes scalable once it's vectorised, because vectors are built from mathematical points and lines rather than pixels (Adobe vector lettering guidance).
Build a supporting glyph set
Even if only the wordmark ships at launch, I still build the full alphabet, numerals, and the diacritics the brand needs. That futureproofs the identity and stops the team reaching for a fallback font later. It also means the brand can set other assets in its own voice, rather than borrowing a style that clashes with the logo.
I run final QA at multiple sizes, checking stem crispness, counter legibility, and silhouette in pure black and pure white. UCL's UK branding guidance on type scaling reinforces the same principle, typography has to stay consistent as size changes, not just look fine at one output level (UCL branding guidance on type scaling).
The practical value of Glyphs is that it turns the logo from a single object into a system component. Once that structure exists, the agency can adapt it for motion, packaging, and digital without redrawing the core letterforms from scratch.
A Lettering Brief Checklist for Brands
A strong brief doesn't ask for “a logo”. It asks for a sequence of decisions the brand is ready to make and fund. If those decisions aren't visible up front, the project drifts into guesswork, and guesswork is expensive in revision time, lost clarity, and weak rollout.
What to approve before final delivery
- Sketch directions: ask for a meaningful spread of early ideas, not a single route.
- Vector concepts: expect more than one clean direction before final selection.
- Small-size legibility tests: make sure the wordmark is checked where it will live.
- Large-format tests: confirm the mark still holds when scaled up for environmental use.
- Do-not-redraw rules: document which features must stay fixed so the identity doesn't drift.
The brand owns voice, values, usage context, and where the mark will appear. The type designer owns stroke contrast, terminal shapes, spacing tolerances, and the way the letters behave under scale. A good brief separates those responsibilities instead of mixing them together.
For teams planning broader implementation, the practicalities of custom fonts setup belong in the same conversation as the logo itself, because a bespoke wordmark often needs a surrounding system to stay useful. If you want a reference for the kind of details that should be captured in a lettering brief, take a look at what to include in a custom font brief.
A logo lettering engagement without those checkpoints is just a hope-and-pray exercise. When the brief names the stages, the deliverables, and the scale tests, the hand-drawn idea can survive all the way into guideline-ready brand assets.
If you need custom handmade type that can live across branding, packaging, advertising, and entertainment without falling apart in production, contact me.