How to Audit Logo Readability Before and After a Redesign

What a readability audit decides

A logo readability audit is a paired, same-condition check of whether the current mark and the proposed redesign remain identifiable, and whether any wordmark remains legible, at the sizes, color reductions, and backgrounds the identity will actually face. The result is a pass, fail, or conditional pass for each intended use, tied to specific artwork—not a preference vote on the canvas.

Readability here covers two related qualities. Legibility is whether letterforms can still be decoded. Recognition is whether the symbol can still be told apart from a generic shape, from noise, or from a competing lockup. A redesign can improve one and weaken the other; the audit records both.

Vector artwork is built from paths: mathematical outlines that describe fills and strokes so the same construction can be scaled without adding pixel steps. Scaling does not preserve readability on its own. Thin strokes, tight counters (the enclosed or partly enclosed interior spaces in letters such as a, e, and o), and dense interior detail collapse at small sizes even when the file remains a true vector. A file extension such as .svg, .ai, or .eps is not proof that the content is vector-only, fully editable, or cleared for reuse; open the file and inspect it.

Starting conditions

Begin only when both versions can be placed in the same tests. If one side exists only as a photograph or a flattened raster, the comparison measures rendering of that image, not the logo’s path construction.

• Current and proposed lockups available as editable vector files, with wordmark, symbol, and combined lockup as separate artboards or layers when those variants exist.
• A written list of intended uses so tests match real constraints: for example a site header, an app or favicon size, embroidery or print, vehicle or signage, and any single-color or reversed application the brand already requires.
• Permission to open, inspect, and—if you will edit—modify the files. Ownership and licence are separate from file format.
• A vector editor that can show outlines, expand or outline strokes, convert type to paths when needed for delivery, and export size-specific previews.
• Agreement on which lockup is in scope. Do not mix a full wordmark on one side with a symbol-only mark on the other unless that substitution is the actual redesign.

How to run the before-and-after audit

Work through the current artwork first, then repeat every step on the redesign with the same sizes, backgrounds, and color reductions. Menu names and export dialogs differ by application and version; treat the checks below as outcomes to confirm in whichever editor you use, and verify commands in that version before relying on a specific control.

1. Inventory the files

Open each file and record what is actually inside it. Note live type versus outlined lettering, stroke-based construction versus closed fills, embedded raster images, clipping masks, opacity tricks, and extra points or overlapping shapes that do not contribute to the visible mark. If lettering is still live type, record the typeface, weight, tracking, and whether optical sizing or manual outlines were used. If it is outlined, you can still audit shapes; you cannot restyle the type without reconstructing it.

2. Normalize the comparison

Place both lockups on identical artboards, aligned to the same optical height rather than to the bounding box if the mark has overshoots (parts of round or pointed letters that extend slightly past the cap height or baseline so they look even). Use the same background colors for both. Lock a small set of test sizes taken from the intended-use list, including the smallest digital size the brand already publishes and a typical print or signage size from that same list. Do not invent a universal pixel or millimetre minimum; the governing size is the smallest size the identity is contracted to appear.

3. Check lettering at every test size

• Counters in a, e, o, g, and similar forms remain open rather than filling in.
• Similar pairs stay distinct: rn versus m, cl versus d, I versus l versus 1, 0 versus O.
• Weight holds together: hairlines do not vanish, and heavy strokes do not fuse with adjacent stems.
• Tracking (letter spacing) still reads as words, not as a texture or as isolated glyphs.
• Diacritics, punctuation, and small caps, if present, remain attached to the right letters and do not become noise.
• Script, condensed, or high-contrast serif forms are checked with extra care at the smallest size, because thin joins and fine serifs fail first.

4. Check the symbol and the lockup

View the mark alone, the wordmark alone, and the combined lockup. Confirm the symbol still has a stable silhouette when interior detail is ignored, that clear space around the lockup is respected in the test frames, and that the relationship between symbol and type does not collapse into one blob. Negative space that forms a letter or a figurative shape should still read after reduction; if it only appears on a large, high-contrast canvas, treat that as a fail for small-size use.

5. Check path geometry that will survive export

• Strokes: if appearance depends on a stroke weight, decide whether final artwork will keep live strokes or be expanded to fills. Live strokes can change visual weight when scaled unless stroke scaling is controlled; expanded fills lock the intended weight. Record which state you audited.
• Minimum feature size: the thinnest bar, serif, or gap in the paths should still be visible at the smallest test size after the file is rendered, not only in outline view.
• Overlaps and compound paths: overlapping fills that rely on stacking order or transparency can shift when flattened or when a printer separates colors. Prefer closed, intended compound paths for counters.
• Open paths and stray points: unused anchors, duplicate shapes, and open contours are delivery defects even if they are invisible at large size.
• Alignment: stems, bowls, and geometric marks should be even optically. Perfect numeric symmetry can still look uneven; the audit is visual at the test sizes.

6. Reduce color and contrast

Repeat the size tests in full color on light and dark backgrounds, then in a single solid color, then reversed (light mark on dark ground). If a one-color or single-ink use is required, that reduction is a required test, not an optional preview. Gradients, shadows, glows, and photographic fills are not a substitute for a readable solid version. Contrast is judged on the actual background colors in the use list; a pass on white does not transfer to a photograph or a mid-tone panel.

7. Record the pair, not a vibe

For each intended use, write the same observations against the current logo and the redesign. A redesign passes the audit when it meets the use-list tests and does not introduce a new failure the current mark did not have at that size—or when a known failure is accepted in writing because that use is being dropped. Preference, novelty, and trend are outside this record.

Check | Current | Redesign | Decision for this use

Counters open at smallest listed size | | |

Confusable letter pairs stay distinct | | |

Symbol silhouette holds without interior detail | | |

Single-color version remains identifiable | | |

Reversed version remains identifiable | | |

Thinnest path feature still visible when rendered | | |

Strokes expanded or scaling behavior documented | | |

No embedded raster standing in for the mark | | |

Live type versus outlines recorded | | |

Clear space and lockup proportions survive reduction | | |

Confirm the result in the delivery files

The audit is complete only when the files you will hand off match the artwork you tested. Outline view and a size-specific export are both required: outline view shows construction; the export shows what a screen, RIP, or printer will actually draw.

• The audited lockups, not an older artboard, are the ones packaged for delivery.
• Type is in the state the recipient needs: live type only when the font is licensed and named; outlines when the recipient will not have the font. Outlines are not reversible without the original type.
• Final marks intended to hold weight at every size are closed fills, or live strokes with the scaling behavior stated in the handoff notes.
• Colors used in the tests match the swatches in the file. Named spot colors, if required by a print specification, are present as named swatches rather than unnamed process approximations—follow the specification you were given; do not invent ink values.
• Export settings produce the preview sizes you judged. Raster previews used for the audit should be generated from the vector, at the target size, without unexplained extra sharpening that hides a thin-stroke failure.
• A one-color and a reversed version exist as real artwork if those uses are in scope.
• Unused layers, hidden tests, and embedded images of the logo are removed so the recipient cannot ship the wrong construction.

Application-specific export checkboxes, SVG decimal precision, PDF compatibility levels, and printer minimum line weights are specification items. Use the values required by the recipient or the print vendor for the job in front of you. This guide does not assign those numbers.

How to read a pass, fail, or conditional pass

Pass: at every size and color reduction on the use list, lettering stays decodable, the symbol stays identifiable, and the delivery files contain the same construction you viewed. Fail: a required use loses the word, fuses the mark, or depends on effects that disappear in one-color or small-size output. Conditional pass: the redesign is acceptable only for a subset of uses (for example a large signage lockup) and a simpler small-size variant still needs to be drawn and audited.

If the current logo fails a use the business still needs, the redesign should be judged on whether it closes that gap. If the current logo already passes and the redesign only matches it, readability is not a reason to change. If evidence is missing—no vector source, no use list, no one-color requirement stated—stop and obtain those conditions rather than scoring a preview.