> For the complete documentation index, see [llms.txt](https://documentation.pushly.com/llms.txt). Markdown versions of documentation pages are available by appending `.md` to page URLs; this page is available as [Markdown](https://documentation.pushly.com/platform/notifications/pushly-ai/brand-voice.md).

# Brand Voice

Brand Voice keeps notifications written with Pushly AI on brand. Describe how your notifications should sound in your own words, like tone, banned words, or a headline length limit. Pushly turns your description into rules and applies them every time Pushly AI writes copy for your domain.

> **Availability.** Brand Voice is part of Pushly AI and is enabled per domain. To turn it on, contact your Pushly account manager.

### What Brand Voice does (and doesn't do)

| Brand Voice **does**                                               | Brand Voice **does not**                          |
| ------------------------------------------------------------------ | ------------------------------------------------- |
| Apply your rules to every AI-generated variant                     | Change copy you write yourself                    |
| Check hard rules (length, banned words, punctuation) automatically | Guarantee tone, which is guidance by nature       |
| Show you exactly which rules it understood before you save         | Silently drop instructions it couldn't understand |
| Let you turn it off for a single variant                           | Use your data to train AI models                  |

### How Brand Voice works

<figure><img src="https://810756845-files.gitbook.io/~/files/v0/b/gitbook-x-prod.appspot.com/o/spaces%2F-Lry9Z27iWOZyQEGAgY7%2Fuploads%2F3NtWif6IKS6ESKzEvMKe%2FScreenshot%202026-10-28%20at%2011.30.18%E2%80%AFAM.png?alt=media&amp;token=c23195bf-20e5-420b-8022-e6e9e0a744ee" alt=""><figcaption></figcaption></figure>

When you save your Brand Voice, Pushly sorts each instruction into one of two types:

#### Guaranteed rules

Rules Pushly can check automatically on every variant. If a variant breaks one, Pushly rewrites it.

Examples: length limits, banned words or phrases, required words, emoji, exclamation marks, all caps, question-style headlines.

#### Guidance

Instructions about tone and style. They shape how Pushly AI writes, but they can't be checked mechanically.

Examples: "confident, conversational tone," "always name the team or player," "emoji on lighter stories."

Anything Pushly couldn't understand is listed back to you so you can reword it. Nothing is silently dropped.

#### What makes a rule guaranteed?

<figure><img src="https://810756845-files.gitbook.io/~/files/v0/b/gitbook-x-prod.appspot.com/o/spaces%2F-Lry9Z27iWOZyQEGAgY7%2Fuploads%2FlCOfXr9rMBZlmpjBSzvM%2FScreenshot%202026-10-28%20at%2011.29.28%E2%80%AFAM.png?alt=media&amp;token=6efbb9e8-85fc-451f-afd0-edb5420b8118" alt=""><figcaption></figcaption></figure>

| Type           | Tip                           | You write                                                | Pushly applies           |
| -------------- | ----------------------------- | -------------------------------------------------------- | ------------------------ |
| **Guaranteed** | Name a number                 | "Keep headlines short" > "Headlines under 55 characters" | Checked on every variant |
| **Guaranteed** | Quote the exact words         | "No clickbait" > "Never use 'you won't believe'"         | Checked on every variant |
| **Guidance**   | The subject changes each time | "Always name the team or player"                         | Shapes the writing       |
| **Guidance**   | The rule has exceptions       | "Emoji on lighter stories, never on breaking news"       | Shapes the writing       |

> **Wording tip:** Words like **must**, **never**, **always** and **under N characters** tend to become guaranteed rules. Words like **prefer** and **aim to** stay as guidance. Write the rules you can't compromise on as firm, specific statements.

### Set up your Brand Voice

You need domain admin access to set up Brand Voice.

1. Go to **Settings > AI > Brand Voice** and click **Set up**.
2. In **Edit your Brand Voice**, describe your writing style in your own words.
3. Click **Generate rules**. Pushly shows how it understood your text: your **Guaranteed** rules, your **Guidance**, and anything it couldn't understand.
4. Review the result. If something isn't right, edit your text and click **Generate rules** again.
5. Click **Save** when the rules match what you meant.

<figure><img src="https://810756845-files.gitbook.io/~/files/v0/b/gitbook-x-prod.appspot.com/o/spaces%2F-Lry9Z27iWOZyQEGAgY7%2Fuploads%2F2US7QK5tNB2L7LZlXMk8%2FScreenshot%202026-10-28%20at%2011.29.05%E2%80%AFAM.png?alt=media&amp;token=e6933cc8-93e0-4a64-a97d-a9466e69b09c" alt=""><figcaption></figcaption></figure>

After you save, the Brand Voice page shows your saved text and how many Guaranteed and Guidance rules it contains.

> **Note:** Pushly saves exactly the rules you reviewed. If you change your text, click **Generate rules** again before saving. If someone else saved a new version while you were editing, you'll be asked to review the updated rules first.

#### Example

**You write:**

> Headlines must be under 55 characters. Never use "you won't believe." No exclamation marks. Always name the team or player. Keep the tone confident and punchy. Emoji on lighter stories, never on breaking news.

**Pushly generates:**

| Guaranteed                    | Guidance                                         |
| ----------------------------- | ------------------------------------------------ |
| Headlines under 55 characters | Name the team or player                          |
| Never use "you won't believe" | Confident, punchy tone                           |
| No exclamation marks          | Emoji on lighter stories, never on breaking news |

### Using Brand Voice when you create a notification

When you generate copy with **Pushly AI** in the notification builder:

* You get **1 to 4 variants**, each written to follow your Brand Voice.
* Each variant has its own **Brand Voice** switch, which is **on by default**. Turn it off to generate that variant without your Brand Voice.
* Each variant has its own **tone**. Choose a preset or write a custom one.

<figure><img src="https://810756845-files.gitbook.io/~/files/v0/b/gitbook-x-prod.appspot.com/o/spaces%2F-Lry9Z27iWOZyQEGAgY7%2Fuploads%2FxvLC6nrChYL2BnuE3BTV%2FScreenshot%202026-10-28%20at%2011.30.28%E2%80%AFAM.png?alt=media&amp;token=15c05375-1e2f-4e07-aa26-acdba2e242e9" alt=""><figcaption></figcaption></figure>

#### When instructions conflict

Pushly AI follows instructions in this order:

1. Stay accurate to the article.
2. Your Brand Voice limits and content rules.
3. The variant's own settings (for example, tone).
4. Your Brand Voice style guidance.
5. Pushly's default writing style.

### Edit or remove your Brand Voice

**Edit:** Go to **Settings > AI > Brand Voice**, click **Edit**, update your text, click **Generate rules**, review, and **Save**.

**Remove:** Click **Edit**, then **More actions > Remove Brand Voice**. Pushly AI goes back to its default writing style immediately.

> **Tip:** To skip your Brand Voice for one notification without removing it, turn off the **Brand Voice** switch on that variant.

### Limits

| Item                                                       | Limit                                                     |
| ---------------------------------------------------------- | --------------------------------------------------------- |
| Brand Voices per domain                                    | 1                                                         |
| Who can set it up                                          | Domain admins                                             |
| Applies to                                                 | Copy generated with Pushly AI in the notification builder |
| Conditional rules (for example, "only for sports stories") | Applied as guidance, not guaranteed                       |

### FAQ

**Does Brand Voice change notifications I write myself?** No. It only applies to copy generated with Pushly AI.

**Does Brand Voice apply to Pushly AI Curator notifications?** Not yet. Curator notifications use the article's own title and description.

**Is my data used to train AI models?** No. Your Brand Voice is used as instructions when Pushly AI writes copy. No client data is used to train models.

**Why did one of my rules end up as guidance?** It probably depends on context, like a rule with exceptions or one that changes with each article. To make it guaranteed, make it specific: use a number or quote the exact words.

**Can I allow a banned word in some cases (for example, inside a quote)?** Not as a guaranteed rule. Move it to guidance with conditional wording, for example "Avoid 'shocking' unless it appears in a direct quote."

**Saving failed. What happened?** Saving needs Pushly to process your text. If that step fails, nothing is saved. Try again.

**Can I have different Brand Voices for different domains?** Yes. Each domain has its own Brand Voice.


---

# Agent Instructions
This documentation is published with GitBook. GitBook is the documentation platform designed so that both humans and AI agents can read, navigate, and reason over technical content effectively. Learn more at gitbook.com.

## Querying This Documentation
If you need additional information that is not directly available in this page, you can query the documentation by asking a question.

Perform an HTTP GET request on the following URL with the `ask` and `goal` query parameters:

```
GET https://documentation.pushly.com/platform/notifications/pushly-ai/brand-voice.md?ask=<question>&goal=<user_goal>
```

`ask` is the immediate question: it should be specific, self-contained, and written in natural language.
`goal` is what the user is ultimately trying to achieve, the reason they need the answer. Sharing it helps GitBook give you a better, more relevant answer. A goal is most helpful when it describes the outcome the user wants rather than restating the question. For example, with `ask=how do I create an API token`, a goal like `automate deployments from our CI pipeline` lets GitBook tailor the answer to that use case.

The response will contain a direct answer to the question and relevant excerpts and sources from the documentation.

Use this mechanism when the answer is not explicitly present in the current page, you need clarification or additional context, or you want to retrieve related documentation sections.
