What Makes a Good Logo? 5 Principles You Can Actually Test

Designers have repeated the same five principles for half a century — simple, memorable, timeless, versatile, appropriate — and they've survived because they describe physics, not fashion: the conditions a logo meets in the real world. The problem is they're usually stated as vibes. Here they are as tests you can run yourself, on any design, before any money changes hands.

The 5 principles and their tests

Principle	What it means	The test you run
Simple	One idea, few elements	Count the elements; describe it in one phrase
Memorable	Sticks after one exposure	The sketch test
Timeless	Not riding a trend	The "which decade?" question
Versatile	Works everywhere	The 16px test + the photocopy test
Appropriate	Fits industry and audience	The lineup test

1. Simple — count the elements

Apple: one shape. Nike: one stroke. Target: two circles. Now count the elements in the design you're evaluating — symbol, name, tagline, container, gradient, shadow. More than three or four distinct elements is a warning sign, because every extra element is something that blurs at small sizes and gets reproduced wrong by someone, somewhere.

Second check: can you describe the logo in one phrase? "A green mermaid in a circle" works. "A swoosh that's also a leaf with the initials inside a hexagon with a gradient" is a committee, not a logo.

2. Memorable — the sketch test

Show the logo to someone for five seconds. An hour later, ask them to sketch it from memory. People reliably sketch Nike, McDonald's and Apple. If your candidate logo comes back as a vague blob, it isn't bad — it's forgettable, which for a logo is the same thing.

Memorability usually comes from one distinctive decision — an unexpected color in your industry, a hidden detail (the FedEx arrow), a strange-but-right shape — not from adding more stuff. Pick one twist and let everything else be quiet.

3. Timeless — the "which decade?" question

Look at the design and ask: could you date it? Heavy gradients and glossy reflections say 2008. Hipster crossed arrows and badges say 2014. If a logo clearly belongs to a year, it will clearly expire. Shell, Mercedes and Coca-Cola have refined the same marks for around a century — refinement is cheap, replacement is not, because a rebrand resets the recognition you paid years to build (see what a logo actually does commercially).

This is also a buying filter: if a portfolio's designs all share this season's trending style, you're buying a 2026 logo, not your logo.

4. Versatile — the 16px test and the photocopy test

The two most brutally practical tests in this guide:

• The 16px test. Shrink the logo to 16×16 pixels — actual favicon size. Is it still recognizable, or is it noise? Every logo eventually lives as a favicon, an app icon and a tiny avatar; the full matrix of sizes it must survive is in the logo size guide.
• The photocopy test. Convert it to pure black and white — no grays, no color. Does it still work? One day your logo will be faxed, embroidered, laser-engraved, stamped, or printed on a one-color invoice. If recognition depends entirely on a gradient, the design fails the cheapest reproduction it will ever meet.

Add two quick variants: drop it on a dark background (is there a reversed version?) and check that a square/icon crop exists. A versatile logo is delivered as a small system in vector format — what that means file-wise is covered in logo file formats.

5. Appropriate — the lineup test

Appropriate doesn't mean literal — Apple sells no fruit. It means the tone fits: a law firm mark shouldn't feel like a candy brand, and a kids' brand shouldn't feel like a bank. The test: place the design in a lineup next to three or four real competitors' logos. You're looking for two things at once — it belongs to the category (a customer would plausibly guess the industry) and it stands apart within it (it isn't the fourth blue shield in the row). Both color and logo type carry most of this signal, and color conventions per industry are real enough that breaking them should be a deliberate choice, not an accident.

Run the whole battery in ten minutes

1. Count the elements (≤4) and describe it in one phrase.
2. Sketch test on one person who hasn't seen it.
3. Ask "which decade is this from?" — the right answer is a shrug.
4. Shrink to 16px; then photocopy-test in pure black and white.
5. Line it up against three competitors.

A design that passes all five isn't guaranteed to be great — but a design that fails two or more will cost you money later, however pretty the presentation slide. Run the battery on every concept you're shown during the design process, and say which test a concept failed when you give feedback; it's the most useful sentence a client can utter.