> ## Documentation Index
> Fetch the complete documentation index at: https://docs.guhan.ai/llms.txt
> Use this file to discover all available pages before exploring further.

# Set up Brand & Voice

> Teach Guhan how your outreach should sound.

Brand & Voice is where you calibrate Guhan's personality. Every message it writes references this page. Get it right once → every future draft is on-brand.

## The five tiles

The Brand & Voice page organizes your settings into five tiles:

<CardGroup cols={2}>
  <Card title="Brand" icon="tag">
    Industry, company one-liner, target customer.
  </Card>

  <Card title="Voice" icon="music">
    Tone, language, phrases you always/never use.
  </Card>

  <Card title="Competitors" icon="users">
    Who you compete with — used for competitor-signal Watchlist Agents + to avoid contacting their employees.
  </Card>

  <Card title="Templates" icon="file-lines">
    Saved message templates you can drop into any sequence step.
  </Card>

  <Card title="Products" icon="box">
    What you sell — each product is a first-class record you can pick per Outreach Agent.
  </Card>
</CardGroup>

## Brand tab

Fields:

* **Industry** — pick from LinkedIn's taxonomy (the same one your prospects' companies use)
* **Company description & value proposition** — 2-3 sentences explaining what you do
* **Customer pain points** — specific problems your customers face; used by the message agent to anchor openers
* **Company website + name** — auto-filled from onboarding

<Tip>
  The more specific your pain points, the sharper Guhan's messaging. "Marketing teams struggle to attribute revenue" beats "Marketing has attribution issues."
</Tip>

## Voice tab

Fields:

* **Preferred tone** — Professional / Casual / Friendly / Direct / Playful / Professional-casual
* **Preferred language** — English (default), Spanish, French, German, etc.
* **Always include** — phrases you want in every message (max 3)
* **Never say** — banned phrases + words that should never appear ("just wanted to reach out", "hope this finds you well", etc.)

Guhan's message agent respects both lists strictly. If you ban a phrase, it won't appear even in follow-ups.

## Competitors tab

Add competitor **LinkedIn company URLs**:

* Used by the "Engaged with competitor" Watchlist Agent signal — finds people who liked/commented on competitor posts
* Used to **auto-exclude competitor employees** from your Watchlist Agent sweeps

Each competitor row lets you add a short **pivot note** — how you position against them ("We're the affordable alternative", "We integrate better"). The message agent uses this to differentiate in outreach.

## Templates tab

Message templates are pre-written **briefs** you can pick when configuring a sequence step. Guhan ships 10 defaults:

* Cold connect · no pitch (LinkedIn invite)
* First message after connect · deliver value (DM / email)
* Signal hook — funding raise (DM / email / WhatsApp)
* Signal hook — hiring (DM / email / WhatsApp)
* Signal hook — post engagement (DM only)
* Follow-up · new value (Day 7) (DM / email / WhatsApp)
* Channel shift · 10-min call ask (DM / email / WhatsApp)
* Break-up · graceful close (DM / email / WhatsApp)
* First email · async version (email only)
* Substantive post comment (LinkedIn post comment only)

You can **clone** any template + customize it, then use the cloned version in sequences. You can also create templates from scratch.

<Note>
  Templates aren't blast text. They're briefs Guhan uses to draft per-prospect messages. So the same template produces different actual wording per person.
</Note>

## Products tab

Every workspace can have multiple products. Each product has:

* **Name** + **short description**
* **Value props** — bullet points of what makes it valuable
* **Ideal use cases** — when it's the right fit
* **Pricing** — optional; used for pricing-forward messaging

When you create an Outreach Agent, you pick which product it's selling. The message agent uses the product's value props + use cases to write messages.

Multiple products = multiple Outreach Agents, one per product.

## Saving and using

Every change on Brand & Voice takes effect on the **next message drafted**. In-flight drafts (queued in the last hour or so) may use the old settings — new sequences from that moment forward use the new settings.

<Warning>
  Changing your tone or banned-phrases list mid-campaign will make messages feel inconsistent to prospects who've received earlier messages. Best practice: change it once, at setup, and keep it stable.
</Warning>

## Related

<CardGroup cols={2}>
  <Card title="Personalization" icon="wand-magic-sparkles" href="/agents/outreach-agents/personalization">
    How Brand & Voice feeds the message agent.
  </Card>

  <Card title="Write message templates" icon="file-lines" href="/brand-and-voice/templates">
    Deep-dive into templates.
  </Card>
</CardGroup>