How to Test a Logo Redesign: A Practical Before-and-After Audit

To test a logo redesign, run the current logo and the proposed one through the same set of checks and compare the results side by side. Inspect how each is built as vector artwork, check the typography, reduce both to the smallest sizes they will actually appear at, try them in one color and reversed, place them in real contexts, and export them in every format the brand needs. The redesign passes when it performs at least as well as the original wherever the brand uses it. Any loss should be a deliberate trade-off, not one you discover after launch.

In this guide, a vector logo means artwork built from mathematically defined paths. A path is a line or curve joined at anchor points. Unlike a grid of pixels, a path can be scaled without losing sharpness. A before-and-after audit is a structured comparison that gives both versions identical tests, so any difference in the results comes from the design and not from how each version was tested.

What to have in place before testing

• Final vector master files for both the current and the redesigned logo, including every variant you plan to use. That means the primary lockup (the fixed arrangement of symbol and wordmark), stacked or horizontal versions, the symbol on its own, and one-color versions.
• A list of the real places the logo will appear, such as website header, social profile image, favicon, app icon, business cards, signage, packaging, merchandise, embroidery and vehicle graphics. Include the smallest and largest size you expect for each.
• Any documented requirements you must meet, such as printer or production specifications, platform guidelines for profile images or icons, and vendor requirements for signage, engraving or embroidery. Treat these as requirements. The checks in the rest of this guide are design recommendations unless a specification says otherwise.
• Typeface information for each wordmark: the font names, the license terms, and whether the text is live or has been converted to outlines.
• Confirmation that you have the right to modify, reproduce and reuse the original artwork, especially if it was created or supplied by a third party.

Don't assume a file is clean vector artwork because of its extension. An AI, EPS, PDF or SVG file can contain embedded pixel images, live text that depends on fonts you may not have, or a mix of both. The extension also tells you nothing about whether you are permitted to reuse or alter the logo.

Step 1: Audit the vector construction

Open both master files and switch to your editor's outline or wireframe view. This view hides fills and strokes and shows only the underlying paths, which exposes construction problems that the normal preview conceals. Most vector applications have this view, but its name, menu location and shortcut differ between applications and versions, so check the documentation for the software you use.

• Embedded or linked raster images: pixel content appears as an image frame or box rather than as editable paths. A vector master should not depend on it.
• Stray and excess anchor points: isolated points left from editing, or too many points along a curve that should be smooth. Both can cause lumpy curves and messy files.
• Open paths: shapes that should be closed but have a gap between their end points.
• Hidden overlapping shapes: objects stacked beneath others that don't show in preview. They can cause problems in single-color production such as cutting, engraving or embroidery, so ask the vendor how they handle overlaps.
• Unexpanded strokes: a stroke is a line whose thickness is an attribute rather than part of the shape. Depending on scaling settings, stroke width may not scale with the logo, which changes its proportions. Converting strokes to filled outlines (often called expanding) locks the shape for delivery files. Keep an editable copy alongside them.
• Shapes with holes: the empty center of a letter such as O (its counter) should be a real hole, created as a compound path, not a white shape covering the fill. To test this, place the logo on a colored background. A fake hole shows up as a white patch.
• Color definitions: every fill should use an intended, defined color value in the color model the output needs, such as spot color, CMYK process color for print, or RGB for screens.

Step 1 is complete when you can confirm the following for each version: every visible element is a vector path, there are no stray points or unintended open paths, counters are true holes, and every color is a defined value. Record the old logo's defects too. Some problems blamed on the old design are actually construction faults, and they will carry into the redesign if parts of the old file are reused.

Step 2: Check the typography

• Live versus outlined text: live text can be edited but only displays correctly when the font is installed. Outlined text displays reliably anywhere but can no longer be retyped. Delivery files are commonly outlined. Keep a live-text source file and record the typeface used.
• Font licensing: confirm that the license covers use in a logo. Terms vary between foundries and licenses, so read the actual terms rather than assuming.
• Kerning: kerning is the spacing between specific pairs of letters. Look at the wordmark both large and small. Uneven gaps stand out at large sizes, and letters can appear to collide at small sizes.
• Customized letterforms: where glyphs have been modified or joined, check in outline view that the joins are clean and the curves are smooth.
• Legibility comparison: read both wordmarks at each size on your application list and note where either becomes hard to read.

Step 3: Run a reduction test

Place the old and new logos at the same height and reduce them step by step until you reach the smallest size on your application list. Don't borrow a generic minimum size from elsewhere. The right minimum depends on the design and the medium, so any size you set should come from your own testing or from a documented production specification.

• Thin lines disappearing or breaking up.
• Counters and small gaps filling in.
• Small text or taglines becoming unreadable.
• Fine details merging into a blob.
• For screens: export raster versions at the target sizes and view them at 100% on the devices people will actually use.
• For print: proof on the same or a comparable process. An office printer gives only a rough approximation of commercial printing, embroidery or engraving.

Step 3 is complete when you have recorded the smallest size at which each version stays clear in each medium. The redesign's figure becomes the candidate minimum size for its usage guidelines.

Step 4: Test color variants

Show both logos in full color, in single-color black, reversed, and in grayscale, on white, on brand background colors, and on representative photographs. A reversed logo (also called a knockout) is a light version placed on a dark background. Check whether the design depends on color differences that vanish in grayscale or in one color. Some production methods, such as spot-color printing and vinyl cutting, cannot reproduce gradients or transparency, so a design that uses them needs a flat alternative. This step is complete when every color variant on your application list exists as a working file and reads clearly on each background you tested.

Step 5: Compare both logos in real contexts

Build realistic mockups for the applications on your list, such as a website header, a circular profile-image crop, a favicon, a business card, signage and packaging. Present the same mockups for both versions in the same order. Check how each logo balances against surrounding content, whether it keeps its clear space (the protected margin that other elements should not enter), and whether its shape fits common containers such as squares and circles without awkward cropping. If the full lockup doesn't fit a container, note whether a symbol-only variant solves it.

Step 6: Check recognition and distinctiveness

• Silhouette check: blur both logos, or view them from a distance, to judge whether the overall shape is still identifiable.
• Competitor lineup: place both versions among logos from the same market and look for confusing similarities.
• Fresh-eyes check: show both versions to people who haven't been involved in the project and ask what they notice and remember.

Small informal checks like these can reveal obvious problems, but they don't measure how a logo will perform in the market, and their results shouldn't be presented as if they do. If the redesign carries significant business risk, consider formal research run by qualified researchers. No evidence on specific research methods was supplied for this guide.

Step 7: Test exports and delivery files

Export every required format from the redesign, and from the old logo too if you are comparing deliverables. Reopen each export in a different application from the one that created it, because problems often only appear outside the original software. Export option names and defaults vary between applications and versions, so confirm the settings in your own software's documentation.

• SVG: open the file in a browser and in a text editor. Look for embedded raster images, live text that depends on fonts, and a viewBox attribute, which lets the graphic scale to its container.
• PDF or EPS for print: confirm that fonts are outlined or embedded, colors use the model the printer requires, and transparency is handled the way the printer's specification asks.
• PNG or other raster files: confirm they are exported at the sizes needed, have a transparent background where required, and show clean edges at 100% view.

For SVG files, a plain text search gives a quick first check. Matches don't always mean there is an error, but each one tells you where to look.

# Flag embedded images, live text and font references
grep -n -E "<image|<text|font-family" logo.svg

# Confirm the root element declares a viewBox
grep -n "viewBox" logo.svg

Step 7 is complete when each exported file opens correctly outside its source application and matches the master artwork in shape and color. It must also meet any documented specification for the place it will be used.

Record the results in a scorecard

Test | What to check | Current logo | Redesign | Pass condition

Vector construction | Paths only, no stray points, true counters, defined colors | Record findings | Record findings | No unresolved defects

Typography | Outlined or licensed fonts, kerning, legibility | Record findings | Record findings | Licensed, even spacing, readable at required sizes

Reduction | Smallest clear size per medium | Record size | Record size | Clear at the smallest size on the application list

Color variants | One color, reversed, grayscale, varied backgrounds | Record findings | Record findings | Every required variant reads clearly

Context | Mockups, clear space, fit in containers | Record findings | Record findings | Works in every priority application

Recognition | Silhouette, competitor lineup, fresh-eyes feedback | Record findings | Record findings | No confusion with competitors; shape stays identifiable

Export | Each format reopened and inspected | Record findings | Record findings | Every file meets its documented specification

Interpret the before-and-after results

Separate construction defects from design limitations. Construction defects, such as stray points, fake counters or unoutlined fonts, can usually be fixed in the file without changing the design. Design limitations, such as detail that fills in at small sizes or a shape that crops badly in a circle, need a design decision. Where the redesign does worse than the original in a high-priority application, you have three choices: revise the design, add a dedicated variant such as a simplified small-size version, or accept the trade-off and record it in the usage guidelines. Proceed when every priority application has a passing result or a documented decision.

Gaps in this guide

This guide doesn't give minimum sizes, clear-space values or other measurements, because no measured data or specifications were supplied. It also doesn't name application-specific commands, which vary by software and version, and it offers no evidence on how well informal recognition checks predict real-world performance. Take measurements from your own testing, take production requirements from your printers, vendors and platforms, and take software settings from the documentation for the version you use.