# Setting Up Alert Rules

> Configure alerts so you’re notified about spikes, risks, and key events as they happen.

> Configure alerts so you’re notified about spikes, risks, and key events as they happen.

While reports give you regular summaries, **alerts** notify you about important changes as they happen (or soon after).

## 10.3.1 Types of Alerts

Common alert types include:

- **Mention spikes** – triggered when mention volume exceeds a threshold within a timeframe.

*Example:* “Alert if mentions double compared to hourly baseline.”
- **Negative sentiment spikes** – triggered when negative mentions exceed a percentage or count.

*Example:* “Alert if negative sentiment > 40% in the last 2 hours.”
- **Keyword-specific alerts** – triggered when certain high-risk or high-interest keywords appear.

*Example:* product recall terms, crisis terms, specific campaign hashtags.
- **Source-specific alerts** – focused on news sites, reviews, or specific platforms.

*Example:* “Alert if any negative review mentions ‘refund’ or ‘fraud’.”
- **Influencer alerts** – triggered when selected authors or high-influence profiles mention your brand.

## 10.3.2 Creating an Alert

1. Go to the **Alerts / Notifications** section.
2. Click **Create Alert** or similar.
3. Configure the alert:

  - **Project / Scope** – select which project(s) the alert should monitor.
  - **Condition / Rule**:

    - Metric: mentions, sentiment, reach, interactions, specific keywords.
    - Threshold: for example, “> 50 mentions in 1 hour” or “negative sentiment > 30% in 24 hours”.
  - **Filters** – platforms, languages, regions, or specific keywords.

  *Example:* only alerts for negative mentions on X in Malaysia.
  - **Delivery** – email recipients (optionally channel integrations if you have them, such as Slack/Teams); immediate vs digest (every 15 minutes, hourly, or daily).
  - **Name & Description** – use clear names like `SL Brand – Negative Sentiment Spike – Facebook`.
4. Save and activate the alert.

## 10.3.3 Alert Tuning & Best Practices

To avoid alert fatigue:

- Start with **few, critical alerts** (for example, crisis-related).
- Use meaningful thresholds (not too sensitive).

Review alerts regularly:

- Disable or adjust alerts that fire too often with low value.
- Tighten keywords or filters to improve relevance.

Combine alerts with dashboards and reports:

- Alerts tell you **when** something needs attention.
- Dashboards and Mention views show you **what** is happening and **why**.