Repic tells you where your brand is weak, and helps you fix it.
Completeness scoring on every component, plain-language weakness alerts, and one-tap fixes.
A score you can argue with.
A number with no reasoning behind it is a horoscope. Every score here names what it looked at and what it would take to move it.
- 01
It reads the kit
Palette, type, logo direction and voice, as they stand right now.
- 02
It reads the output
What you have actually published, not what the kit says you would.
- 03
It scores the gap
The distance between those two is the whole assessment.
- 04
It says what to change
Ranked by how much it would move, so the first fix is the one worth doing.
Statuses that tell the truth, per component.
Weakness alerts, with a fix attached.
One line isn't enough for the engines to write to. Add 2-3 concrete struggles your audience would recognize.
✦ Improve this fieldStaleness, tracked: change a source field and Repic knows which parts of your kit are now out of date. Weekly digest, optional: a summary of weaknesses by email, only if you turn it on.
Independent scores, honest roll-up.
3 min read · 6 sections
Why a brand needs a health check, not a launch date
Brands do not fail at launch. They decay. A field goes stale, a new offer never makes it into the persona, an audience description written for last year keeps producing copy for last year audience, and nobody notices because nothing broke. The output just gets quietly less good, and there is no error message for that.
The Quality Engine exists to make that decay visible. Every persona field and every kit module is saved and scored independently on completeness, specificity and consistency. Those component scores roll up into the brand health figure on your Studio overview, so there is no single mystery number you cannot decompose.
Keep reading
The scores are not the product; the alerts are. An Underspecified badge on a field is accompanied by a plain-language statement of what is missing and a one-tap path to fix it. And because Repic knows which generated work depends on which source field, editing a field tells you what is now out of date instead of leaving you to guess.
Step by step, with nothing skipped
Everything is scored independently
Per persona field, per kit module: completeness, specificity, consistency. No aggregate that hides which part is weak.
Statuses tell the truth
Healthy means ready to build on. Underspecified means it works but generation quality is capped. Insufficient means there is not enough to generate from at all.
Every alert carries a fix
The alert names what is missing in plain language, and Improve this field takes you straight to it.
Staleness is tracked, not guessed
Change a source field and Repic knows which parts of your kit and which past runs are now out of date. An optional weekly digest summarises it by email, only if you turn it on.
An alert that is actually actionable
Before
Status: Underspecified. Audience, "What they struggle with".
After
One line is not enough for the engines to write to. Add two or three concrete struggles your audience would recognise. [Improve this field]
The difference between a score and an alert is that a score tells you that you are at 62% and an alert tells you which sentence to write next.
What this does not do
It does not grade your taste
It measures whether the engines have enough to work with, not whether your brand is good. Those are different questions and only one of them is measurable.
It does not fix things on its own
It shows you what is weak and takes you to it. Repic does not rewrite your brand while you are not looking.
It is not an SEO or engagement score
This is about the completeness and internal consistency of your brand profile, not about how a specific post will perform.
Straight answers
What is the difference between Underspecified and Insufficient?
Underspecified means the field works but is thin, so generation quality is capped by it. Insufficient means there is not enough there to generate from at all, and the alert says exactly what is missing. Both are shown with the same badge components you see inside the app, never lookalikes.
Does a high score guarantee good output?
No, and it does not claim to. A high score means the engines have specific, consistent, complete input. That removes the most common cause of generic output; it does not make every run excellent.
If I change one field, does my whole brand rescore?
No. Only what depends on that field rescores. The rest of your brand stays untouched, which is what makes editing safe enough to do often.
Will it email me about problems?
Only if you turn the weekly digest on. It is off by default, because an unrequested weekly email about your brand health is a notification, not a service.
Read next
- the Branding engine that writes the fields this scores
- asking the Assistant what is weak in your brand
- Repic for recruiters and talent teams, whose employer brand goes stale after every repositioning
- step five of the journey, staying healthy
- what an AI brand agent is, and what it is not
- on-brand content, defined so you can test for it
- the brand glossary, 20 terms defined
- why AI-generated content drifts off brand
- the on-brand content checklist
Free, no signup
Not ready to sign up? Run the brand consistency checker instead.
Paste two or three posts and get a drift score with the evidence: sentence rhythm, who you address, register, filler phrases and palette contrast.