What You’re About to Throw Away

A logo isn’t one thing you keep or replace

The framing that causes most of the damage is treating the mark as a single object with a single decision attached: keep it, refresh it, or start over.

A logo is a set of decisions bundled together. Letterform character. Weight. Shape language. Proportion and spacing. Color relationships. Texture or the absence of it. How it behaves next to photography and type. Each one of those is doing a different amount of work in your audience's head, and the amounts are wildly uneven.

Figure 1 · One mark, six decisions, uneven loads

Which means the useful question isn't “should we redesign the logo.” It is “which of these decisions is load-bearing.” You can change everything that isn't, sometimes dramatically, without anybody feeling a loss.

Familiarity to the team is not equity in the market

Two mistakes sit on either side of this, and teams make them in the same meeting.

Killing something that’s carrying meaning

The people closest to a brand get tired of it years before the market does. What looks exhausted from inside often reads as stable and familiar from outside, because customers see it in passing, a few seconds at a time, in a context where they aren't evaluating it at all.

Keeping something because it’s familiar to the room

The reverse happens just as often. An element survives every redesign because it is sacred internally, tied to a founder or a legacy or an argument nobody wants to reopen, while carrying no weight whatsoever with the audience. It gets protected, the rest of the identity gets built around it, and it constrains the work for another decade.

Figure 2 · The two mistakes, and the two safe cells

Both mistakes come from the same source. Internal attachment and audience association feel identical from inside the company, and they are routinely opposites.

What to measure: association, not preference

The instinct in a redesign is to ask which version people like. Preference is the least useful thing you can measure here, because it tells you what people would choose in a moment that never happens. Nobody in your market will ever be shown your old mark and your new mark side by side and asked to pick.

Association is the measure that predicts anything. What does this shape, this weight, this color relationship connect to in someone's head? Established or scrappy. Expensive or accessible. Careful or fast. Regional or national. A mark can be handsome, clear, well crafted, and associate with the wrong category or the wrong price tier, and that is the failure that shows up as a slow drift in the year after launch rather than an immediate outcry.

Recognition is a separate thing again. People can recognize a mark and connect nothing to it, which is a finding worth having, because it means you are free.

How to find out before you design

The sequence looks a lot like the measured brand audit, pointed at the mark instead of the material.

1. Take the current mark apart. Pull out its elements as raw material: the letterform character, the shape, the proportion, the color relationships, any texture. Not the assembled logo. The parts.
2. Put those elements into a wider set. Test them inside a range of two dozen to four dozen elements that includes alternatives, adjacent directions, and material with no relationship to your brand at all.
3. Don't announce which ones are yours. The moment an audience knows they are being asked about a brand's current logo, you have introduced a different question and a different set of manners.
4. Segment the audience. Employees and long-time customers versus prospects and people new to the category. That split is where a redesign's real risk lives, and you want that fork visible before you commit rather than averaged into a middle that describes nobody.
5. Read the clusters against the intent. Where responses cluster on elements you were planning to keep, you have equity, and you now know what it is made of. Where your intended keeps produce nothing, they are old rather than meaningful. Where an element you had already written off shows up strongly, that is the finding worth the whole exercise.

Figure 3 · Preference test or association test

Results come back in under 48 hours, which means this fits before the design work rather than instead of it.

Redesign as a risk decision

Once the elements are separated into load-bearing and not, the conversation with the client changes shape entirely.

You are no longer arguing about whether the new mark is better, which is an argument between two tastes with no referee. You are presenting a map of what the audience is holding and proposing to change everything else. The elements carrying association stay, possibly in a completely new form, since carrying meaning forward isn't the same as keeping a shape frozen. The elements carrying nothing are free.

That map also settles the internal fight, because it is very hard to defend killing something the audience clearly connects to, and very hard to defend protecting something they clearly don't. Anyone who wants to override it needs evidence of their own.

The redesign still requires judgment, craft, and a creative director willing to make a call. It just stops requiring a guess about which parts of the current brand are load-bearing, which is the guess that produces the redesigns nobody can undo.

Frequently asked questions