Wordmark or symbol? How to decide before you brief a designer

Founders usually arrive at this decision having already made it. “We need a logo and an icon” comes up in a first call about as often as “we need a logo”, and almost nobody says why. The symbol gets assumed the way a favicon gets assumed.

Slow down for one afternoon. The two options cost different amounts and break in different ways, and a symbol is a second asset with its own life. It has to be designed, tested, defended in review, and then kept working on every surface the company adds for the next decade.

Taste doesn’t decide this. An inventory does.

The three configurations

Wordmark only. The name, set in a typeface that has been drawn or modified for the purpose. Nothing else. Most B2B software that doesn’t ship a mobile app can live here permanently.

Wordmark plus symbol, used as a lockup. The two travel together in marketing, and the symbol is allowed out alone where a wordmark won’t fit. This is the most common setup, and the one most often shipped incomplete: the lockup gets drawn, the rules for splitting it don’t.

Symbol-first. The symbol carries recognition on its own, and the name appears beside it only when an introduction is needed. Every company wants to be here. It’s earned by repetition at a volume most early-stage products won’t reach for years, and designing for it before you have that exposure buys nothing.

The inventory that actually decides it

Write down every surface the mark has to appear on in the next two years. Yours, not the ones you admire on other companies’ brand pages.

The list that matters is short, and every item on it is a square:

• Browser tab. 16px, and the first place a detailed symbol dies.
• Mobile app icon. Only if you have an app or will. This is the strongest argument for a symbol, because the slot is square, mandatory, and a wordmark in it is illegible.
• Social avatar. Square, usually rendered as a circle, so the corners aren’t yours.
• Product UI. A collapsed sidebar, a 32px top bar, a loading state.
• Partner directories and marketplace tiles. Someone else’s grid and aspect ratio, no negotiation.
• Anything physical. One-colour print, embroidery, laser etching. Your gradient doesn’t survive any of them.

Count the squares. If the honest answer is “the favicon, and that’s it”, a wordmark plus a carefully made favicon built from its first letter will serve you, and a real share of the identity budget goes to work that moves revenue instead. If the list has an app icon and a marketplace tile on it, you need a real symbol, and you needed it before the launch date.

That’s the whole decision. The rest is execution.

The five tests

Whatever comes back from the designer, run these before you approve it. Ten minutes, and they separate a mark that ships from one that gets quietly redrawn eight months later.

1. The 16px test. Render the symbol as a favicon, at actual size, in a real browser tab next to your other tabs. Inner detail, thin strokes and letterforms inside the shape vanish here. If you can’t tell it apart from the tab beside it, it fails, however good it looked on the presentation slide.

2. The circle crop. Put it in a circular avatar frame. Anything living in the corners is gone. A symbol drawn on a square canvas without this check loses a limb.

3. One colour. Flatten it to solid black, then to solid white on a dark field. Gradients, overlaps and transparency that carried the shape now carry nothing. If the silhouette doesn’t read on its own, colour was holding it up.

4. The dark screen. Drop it on your product’s darkest surface. A mark drawn on white often needs a different weight, or an inverted version with its own rules, and discovering that during the dark-mode sprint is the expensive way.

5. The blur test. Blur it until the detail is gone and look at what’s left. Recognition at a distance is silhouette. A symbol tends to pass this on the first try or never.

A wordmark gets a shorter version of the list: a minimum legible size written down, a one-colour file, and a version with enough letter-spacing to survive small rendering.

Where symbols go wrong

The meaningless swoosh. A shape chosen because the deck needed a shape. It survives review and is forgotten in the world. If the rationale takes a paragraph, the symbol isn’t doing its job, which is to be recognised without explanation.

The one-angle mark. Beautiful in the hero image, illegible at 16px, mutilated by the circle crop. Almost always a sign it was designed at presentation size and never taken down.

The lockup with no rules. Wordmark and symbol are drawn together and nobody writes down the minimum size, the clear space or, the one that matters, when the symbol may go out alone. Six months later there are four versions in circulation, each made by a different person solving the same problem.

The missing plain version. There’s no one-colour file, so the first time a conference programme asks for mono artwork, someone recolours it by hand at 11pm.

To see these decisions finished, in identities you already know, the annotated set is in 12 SaaS brand identity examples. Each one is scored at 16px, in dark and light, and as tokens.

What this is not

It isn’t the same question as “what is our brand”. A logo is one artefact inside an identity, and mixing the two up is the most expensive mistake in this area. That one has its own post: brand identity vs visual identity vs logo.

You also don’t have to answer it alone in advance. A competent engagement starts with the surface inventory and lets it decide, which is how our logo design process opens. Arriving with the inventory already written down still saves a week of circling.

The decision, in one page

1. List the square surfaces you’ll actually need in two years.
2. Zero or one of them, and no app: wordmark, plus a favicon derived from it.
3. An app icon, a marketplace tile or a compact product UI: wordmark plus symbol, and write the rules for splitting the lockup the day you approve it.
4. Symbol-first: not yet. Revisit when people recognise the mark without the name.
5. Run the five tests before approval, not after.
6. Whatever you ship, make the one-colour version at the same time. Twenty minutes now, an emergency later.

Already have a mark and not sure which of these it fails? The brand audit checklist asks the same questions as a twenty-item self-assessment you can run without a designer. If the inventory says you need a symbol and an app icon designed together, that’s a brand identity engagement, priced on the page.

FAQ

Do we need a symbol if we don’t have a mobile app? Usually not yet. Without an app icon, the only mandatory square is the favicon, and a favicon can be built from the first letter of a wordmark. Spend the difference on the parts of the identity customers actually read.

Can a wordmark work as an app icon? Only if the name is very short, and even then it’s a compromise. The slot is small and square, which is exactly what a wordmark handles worst. If an app is on the roadmap, design the symbol with the identity instead of bolting it on before launch.

What is a logo lockup? The defined arrangement of wordmark and symbol (relative size, spacing, alignment) together with the rules for when each may be used alone. The arrangement without the rules is the incomplete version most teams ship.

How do we know if our symbol is any good? Run the five tests above: 16px, circle crop, one colour, dark background, blurred silhouette. They’re objective, they take ten minutes, and a mark that passes all five won’t need redrawing when the next surface appears.

Should the symbol mean something? It should be recognisable before it’s meaningful. Meaning that needs a paragraph of explanation isn’t doing the job. A shape people can pick out of a row of tabs is.