Mentova continuously monitors your visibility signals and writes an alert the moment a real problem shows up — or clears. An alert is not just a log entry: it describes a problem ("you dropped off a query", "a competitor overtook you"...) and stays open across analyses for as long as the problem persists.


The 12 detectors

Every campaign triggers a detection pass on the signals it just measured. A second, daily pass also watches your account independently of any campaign.

Available on every plan, including FREE:

  • Your visibility is dropping — your mention rate falls between two analyses.
  • New competitor in AI answers — a competing brand now appears in AI responses.
  • Your credits are running low — you're almost out of research credits this month.
  • No recent analysis — you haven't run a campaign in a while.
  • A draft is waiting to be published — a Mentova-generated draft was never published.

From the STARTER plan, seven more detectors kick in:

  • You disappeared from a query — a model used to cite you on a specific query and no longer does.
  • You appeared on a new query — the positive counterpart: a model now cites you where it didn't before.
  • A competitor overtook you — the share-of-voice gap with a competitor is widening.
  • One AI cites you a lot less — one specific model cites you well below the average of your tracked models.
  • Tone is worsening — how AI answers talk about you is becoming less favorable.
  • A source cites your competitors, not you — an influential site mentions several competitors but never you.
  • One of your markets is falling behind — if you track several markets, your mention rate in one of them sits well below your all-markets average (measured on at least 10 answers in that market).

On FREE, the "What is watched" column still names the paid detectors, locked — without ever showing a number they aren't allowed to measure — a clear invitation to upgrade, never a made-up count.


The Alerts page

The Alerts page, reachable from the navigation menu, fits on one screen:

  • A summary line and four figures — To handle, Under watch, Resolved this month and Active detectors ("9 / 12"). Clicking one of the first three opens the matching tab; these figures always count the total, whichever tab is showing.
  • "Handle first" — the highest-priority open alert, with a single button to act on it (launch a campaign, view the competitor, create content...). It is not repeated in the list.
  • "Your alerts", in three tabs: To handle (open problems, sorted by priority), Under watch (alerts you snoozed for 7 or 30 days) and Resolved this month (problems that cleared on their own, so you also see what's working, not only what's wrong). Each row shows the problem, its scope with a logo (competitor, AI, site, question or market), the sentence with its numbers, its age and the ⋯ menu (snooze 7 or 30 days, mark as handled, dismiss, reopen). Ten rows show, then "See the N others".
  • "What is watched", on the right (below the list on a small screen) — the 12 detectors by family (Visibility, Competitors, Questions, Image, Account). Each one is active (with "+ email" when the recap email covers it), off because of your settings, or locked with the name of the plan that unlocks it — never showing a number it isn't allowed to measure. A link leads to Settings > Notifications.

An alert that persists across analyses stays the SAME alert — it reads "for analyses" rather than duplicating. Dismissing an alert keeps it gone until the problem clears and comes back.


Per-detector settings (Settings > Notifications)

Each detector is configured independently from Settings > Notifications, on one row:

  • The "Email" checkbox — the alert goes into the recap email.
  • The ⋯ menu — the sensitivity (Low / Medium / High: how marked the shift needs to be before it fires) and "Also send to a webhook" (locked below the Pro plan).
  • The switch — on / off. A detector that is on always shows in the app.

The Account detectors (credits running low, no recent analysis, unpublished draft) are usage reminders: they stay in the app only, with no email checkbox or webhook.

These settings are fully savable on the FREE plan: nothing goes out until your plan allows it, but your preference is saved and applies automatically the day you upgrade — nothing to reconfigure.

Recap email (STARTER and above)

A recap email bundles up to 3 not-yet-notified alerts per detection pass. It reuses the exact same copy as the app.

Alert webhooks (PRO and above)

A webhook (Zapier-compatible) receives every alert as JSON the moment it's detected.

Go to the Publications page (Automation › Publications), where all your connectors live, open the tile for the relevant service, then tick "Send alerts to this connector" on the webhook connection's row.

The checkbox only appears on webhook connectors: WordPress, Webflow and GitHub publish content, they cannot receive alerts.

Until you tick it, the connector receives only your content publications — an existing publishing Zap never starts receiving alerts by accident. Then choose which alerts go out by webhook, detector by detector, in Settings > Notifications.

Weekly digest

Independently of the detectors, a weekly visibility summary is sent every Sunday at 8:00 UTC if you enable it in Settings > Notifications.


Dashboard

The number of open alerts also shows in the header of every page, with a link to the Alerts page.