Alerts

Alert types

Inception Agents provides four categories of alerts to keep you informed about changes that affect your AI visibility and optimization performance.

Visibility changes

Triggered when your AI Share of Voice shifts beyond configured thresholds:

• SoV drop — Your Share of Voice on one or more platforms decreases by more than the configured threshold (default: 10% relative change).
• SoV increase — Your Share of Voice increases significantly, indicating a positive trend worth noting.
• Competitor movement — A tracked competitor’s SoV changes significantly relative to yours.
• Category shift — Your ranking within your configured industry category changes.

Traffic anomalies

Triggered when agent traffic patterns deviate from established baselines:

• Traffic spike — Agent visits exceed the rolling 7-day average by more than the configured multiplier (default: 3x).
• Traffic drop — Agent visits fall below the rolling 7-day average by more than the configured threshold (default: 50%).
• New agent detected — An agent not previously seen begins visiting your site.
• Unusual crawl pattern — An agent accesses an unusual number of pages or visits at an atypical frequency.

Optimization events

Triggered by changes in the optimization pipeline:

• New winning variant — The learning engine has identified a new best-performing content variant for a content chunk and agent combination.
• Honesty flag triggered — The Honesty Engine has flagged content for review. Flagged content is held from serving until resolved.
• Variant performance shift — A previously winning variant has been overtaken by another, indicating a change in agent preferences.
• Content regeneration — New variants have been generated after a site crawl detected content changes.

Integration health

Triggered by issues with your integration infrastructure:

• API errors — The Inception Agents endpoint is returning errors for agent requests.
• Connection issues — The integration (middleware, worker, or DNS proxy) is not reaching the Inception Agents API.
• Crawl failures — Scheduled site crawls are failing or completing with errors.
• Certificate issues — SSL/TLS certificate problems detected on the integration endpoint.

Channels

Alerts can be delivered through three channels. Each alert rule can be configured to use one or more channels simultaneously.

Email

Alert notifications sent to one or more email addresses. Includes a summary of the alert, relevant metrics, and a direct link to the affected dashboard view.

Configure email recipients in Settings > Alerts > Email.

Slack webhook

Real-time alert notifications sent to a Slack channel via incoming webhook. Messages include formatted alert details with links to the dashboard.

Configure your Slack webhook URL in Settings > Alerts > Slack. You can configure separate webhooks for different alert categories.

Dashboard notifications

In-app notifications visible in the dashboard notification center (bell icon). All alerts appear here by default regardless of other channel configuration.

Dashboard notifications are marked as read when viewed and archived after 30 days.

Configuration

Threshold settings

Each alert type has configurable thresholds that determine when it fires:

Thresholds are configured per alert rule in Settings > Alerts > Rules.

Frequency

Control how often alerts are delivered to avoid notification fatigue:

• Real-time — Alert is sent immediately when triggered. Best for integration health issues and honesty flags.
• Daily digest — All triggered alerts for the day are bundled into a single notification, sent at your configured time (default: 09:00 UTC). Best for visibility changes and optimization events.
• Weekly summary — A comprehensive weekly report of all alert activity, sent at your configured day and time (default: Monday 09:00 UTC). Best for trend monitoring and executive reporting.

Each alert rule can have its own frequency setting. Different channels can receive different frequencies for the same alert.

Default alerts

The following alerts are enabled out of the box when you create your account:

Default alerts use the account owner’s email address. Modify defaults or add channels in Settings > Alerts > Rules.

Custom alerts

Create alert rules tailored to your specific monitoring needs:

1. Navigate to Settings > Alerts > Rules > Create Rule.
2. Select the alert type (visibility, traffic, optimization, or integration).
3. Configure the trigger condition (threshold, comparison, or event).
4. Choose delivery channels (email, Slack, dashboard).
5. Set frequency (real-time, daily digest, weekly summary).
6. Optionally set a scope — limit the alert to specific platforms, pages, or competitors.
7. Save the rule.

Example custom rules

• Alert when ChatGPT SoV drops more than 5% relative to the previous week, delivered via Slack in real-time.
• Alert when a specific competitor’s SoV exceeds yours on any platform, delivered via email as a daily digest.
• Alert when agent traffic to your pricing page increases by 2x or more, delivered via Slack in real-time.
• Alert when any content variant is flagged by the Honesty Engine, delivered via email and Slack in real-time.

Alert history

View a complete log of past alerts in Analytics > Alerts > History:

• Chronological log — All triggered alerts listed by timestamp.
• Filter by type — Show only visibility, traffic, optimization, or integration alerts.
• Filter by status — Show acknowledged, unacknowledged, or resolved alerts.
• Filter by date range — View alerts from a specific time period.
• Alert details — Click any alert to see the full context: what triggered it, the metric values at the time, and any actions taken.

Alert history is retained for 12 months.

Best practices

Recommended alert configurations for common use cases:

Early-stage monitoring

When you first integrate, focus on integration health and baseline traffic:

• Enable real-time alerts for all integration health issues.
• Set traffic anomaly thresholds conservatively (allow 2-4 weeks to establish baselines).
• Use daily digest for visibility changes until you have stable SoV data.

Active optimization

When you are actively testing and optimizing content:

• Enable real-time Slack alerts for honesty flags — resolve these quickly to avoid serving stale content.
• Set daily digest for new winning variants to stay informed about optimization progress.
• Monitor competitor SoV weekly to track relative positioning.

Steady-state operations

Once your optimization is mature and stable:

• Shift most alerts to daily digest or weekly summary to reduce noise.
• Keep integration health alerts at real-time.
• Set tighter SoV thresholds (5% instead of 10%) to catch smaller shifts early.
• Enable competitor alerts for new entrants in your category.