Brand Settings

Brand Settings holds the defaults and saved building blocks your whole team shares for a brand, so every study starts from the same foundation.

Set it once

Everything on this page exists so a study never starts from a blank box. The walkthrough below goes section by section — it writes the Pulse defaults, adds a competitor, saves a reusable context, and previews an already-declared panel dimension of your own — and then opens Pulse, where a brand's saved setup is waiting. Play it, or step through the five chapters.

What you can save

• Defaults — the baseline brand and product context applied to new studies.
• Competitive landscape — the competitors you test against, with their positioning, audience, strengths, and weaknesses.
• Saved contexts — named context rules that appear as presets in Pulse and reruns.
• Creative copy library — reusable copy, including drafts carried over from Compass.
• Saved segments — audience segments saved from your results, ready to target in future studies.
• Persona attributes — your own dimensions on the panel, like how familiar each person is with your brand.

Persona attributes

The panel is shared by every brand, and persona attributes let you annotate it with something only you care about — how familiar each person is with you, for instance. Your values are never visible to another brand. Choose whether an attribute is shown to the model when it evaluates your work, or kept aside purely for slicing results, and whether a Pulse can be aimed at it.

An attribute can also replace one of the persona's own fields for your brand — their occupation, income bucket or region — instead of sitting alongside it. Tick "Override a built-in profile field" and pick the field: your value is what the model reads and what your result segments are grouped by, while every other brand still sees the panel's original value. An override is always shown to the model and always targetable, so visibility and targeting are fixed for it.

An override changes who a segment describes, not who gets sampled: the geo and age filters you aim a Pulse with still match the panel's original profile. And if you override age without also overriding the age bracket, the two can disagree inside the same persona.

You describe the panel in rates: "among the midwest, 50% are familiar or better". You do not put the rules in any order, and there is no first-match-wins — every rate you state is solved for together, so all of them come out true at the same time. That is the point, but it means the result of two overlapping rates is arithmetic rather than something you can read off the list.

Hold men to 10% while the midwest sits at 50% and midwest women land well above 50%. Nothing has gone wrong — that is the only way both numbers can be true. Always Preview before you save.

Preview before you save

• Preview shows what each rate actually achieves against the real panel, condition by condition, without saving anything.
• It also shows how far each condition moved since your last save — including conditions you did not touch, because the fit rebalances around whatever you changed. Movements are given in people as well as percentages, so you can see the size of what you just changed.
• Below that, Preview lists every group your conditions carve out — including the ones you never wrote a rule for, like "Midwest, not Men". This is where the surprises show up: your stated rates can both be met exactly while a group you never named sits far above or below them.
• Sometimes rates cannot all hold at once. That is not always a contradiction: a cohort can simply be too small to reach the number. Preview marks which ones missed and by how much, and you can choose to save the closest fit anyway.
• A condition that matches nobody at all still shows what it achieved — 0% — but its miss reads "no personas match" instead of a shortfall. Usually the condition names a value no one on the panel carries: a typo, or a value that has since gone away.

"No personas match" is not a rate that missed. There was nobody to apply the rule to, so the number you typed had no effect at all — including if you typed 0%, where nothing else would flag it. Saving is refused until you fix the condition or accept the fit deliberately.

Because groups overlap, the people-counts are given per condition and cannot be added together into a panel total.

Pinning exact values

Rules describe the panel in aggregate. When you need particular people set to particular values, download the CSV template, edit it, and upload it back. It arrives with everyone's current value already filled in, so an untouched round trip changes nothing. A blank cell means "leave this one alone", and a dash puts that person back under the rules. Pinned values win over the rules.

Uploads are all or nothing. Your file is checked first and any bad rows are listed with the line number your spreadsheet shows, so you can fix them before anything is written.

An attribute's key is permanent once saved. It is part of how values are worked out, and renaming it would strand every value you have uploaded against it.

Sharing and permissions

Everyone working in a brand can see its settings, and brand owners and admins can edit them. Switch the active brand from the brand selector in the top bar to manage a different brand's defaults.

Sharing Brand Settings with someone without an account

Signed out, /brand-settings is a public page rather than a sign-in prompt. It plays the brand setup walkthrough: one brand set up end to end — Pulse defaults, a competitor, a saved context, a previewed panel dimension — and then cuts to the Pulse where all four are already waiting, which is the argument for filling any of it in.

What a signed-out visitor sees at /brand-settings. Set up a brand. opens account creation in place; once they finish, the same URL becomes the settings sections described above.

Setting up competitors and saved contexts once here means every new Pulse and rerun picks them up automatically.