Brand Identity

A brand is the perception sitting in customers' heads, which means a company cannot edit it directly. The only lever is deciding what to show repeatedly. For a long time that lever had no name, so asking "what is our brand?" always slid into a conversation about the logo.

Pagedeck went through this. When Grace asked what the product should feel like, Noa answered with colors and typefaces, Ravi answered with ad copy, and Sofia answered with the words customers actually used. None of the three was wrong, and none of the three added up to one brand.

In that state every decision turns into an argument about taste. Three separate meetings went by choosing between two lines of ad copy without reaching a conclusion, and the tie eventually broke toward whoever spoke loudest or sat highest. The same thing happened again on the careers page and again on the pricing page.

Brand identity pulls that argument out of the individual decisions and moves it one level up. Settle in advance who you're talking to, what you promise them, and what character you speak in, and the decisions underneath stop being about preference and start being about fit. If the brand is the output, the identity is the input the company supplies.

It works by fixing a few sentences first and deriving everything visible from them. Four layers, top down.

1. Audience — who the company is talking to. Set it as wide as "anyone who needs documents" and the three layers below have nothing to work with. Narrow it to "the person who keeps meeting notes at a company of 20 to 100 people" and the voice and examples start falling out on their own.
2. Promise — one sentence about what changes for that person. Outcomes, not features. Not "real-time collaborative editing" but "the notes are finished before the meeting is."
3. Character — three or four adjectives describing the brand as if it were a person. One test applies here: the opposite has to be a real option. No company would choose the opposite of "friendly," so it sorts nothing. "Informal" splits against "formal," so it does work.
4. Expression — the three above, translated into things people see and hear. Naming, the logo, the tagline and tone and manner all live in this layer.

The whole point of the order is that each layer comes out of the one above it. If Pagedeck fixes "unfussy" as part of its character, the logo skips ornament, interface messages stay under a sentence, and the pricing page carries no asterisked footnotes. Three decisions that never consulted each other land in the same direction.

You can check the work by reading it backwards. Show someone only the finished logo and copy, ask what kind of brand it seems like, and if the adjectives you wrote come back, it worked. "I couldn't say" means the expression layer isn't carrying anything from above.

It gets pulled into the room less for making something new than for choosing between two options.

Say Ravi brings in two lines of ad copy. One is "boost productivity by 300 percent." The other is "the notes are finished before the meeting is." Argued as taste, that goes nowhere. With "never overstates" written into the character, the first one drops out on the spot. Decisions speed up not because the standard is clever but because the standard existed before the meeting started.

The recurring places look roughly like this.

1. Ad copy and landing pages — which claims are available to the company at all.
2. Job postings — who fits here, which follows from the character.
3. Pricing pages and help docs — whether awkward facts get buried or stated plainly.
4. Support replies — how far to apologize and in what register.
5. In-product copy — down to what an empty screen says.

An identity by itself doesn't make those places line up. Sofia isn't going to recall a list of adjectives every time she builds a proposal. That's why the decisions get transcribed into a brand guideline. The identity holds why; the guideline holds what to do about it.

When the business changes, the identity is the first thing to revisit. Rebranding often stops at a new logo, but swapping the expression while leaving the audience and promise untouched moves nothing on the customer side.

"This doesn't feel like us. Are we a brand that talks this way?"

An objection grounded in a written standard rather than personal taste.

"Let's settle the identity before we look at logo concepts."

Choosing concepts without a standard means choosing by preference.

"'Unfussy' is in our character list, so let's drop the badge on this screen."

A written adjective applied directly to a screen-level decision.

Pagedeck spent half a day on this right after its Series A. Grace got everyone in a room and Mia, Noa, Ravi and Sofia settled one line for the audience, one sentence for the promise, and four adjectives. The longest part wasn't finding candidate adjectives. It was throwing out the ones whose opposite nobody would pick — "innovative" and "trustworthy" went that way.

The return on that half day showed up later. Copy review went from three meetings to one.

The most common misreading treats the identity as the logo and the colors. That's the part visible from outside, so the assumption is natural. But the logo and the colors are products of the fourth layer, and the audience, promise and character that generated them live only in a document. Left uncorrected, this belief leads a company to commission a new logo and conclude it changed its identity. With the promise and the language unchanged, customer perception doesn't shift at all, and the only durable result is the invoice.

The other one is assuming that what the company decides is what gets received. It's written down, so it should land that way. But the identity is the input the company supplies and the brand is what customers ended up holding, and the two come apart often. If "expert" sits in the character list while customers only ever describe the product as easy to use, the thing to fix isn't the customers but what got shown to them over and over.