Changelog

AI visibility leaves beta

• AddedGoogle AI Overviews joins ChatGPT, Gemini and Perplexity as a tracked answer engine.
• AddedCited sources are now recorded per answer, so you can see what an assistant read before deciding.
• ImprovedVisibility score is weighted by assistant reach rather than counted flat.

• FixedGrid scans at 11×11 could time out on service areas wider than 4km.
• FixedCroatian diacritics were stripped from generated replies in rare cases.
• ImprovedReply drafts now avoid repeating the reviewer's exact wording back at them.

Confidence gating for auto-replies

• AddedReviews whose sentiment confidence falls below 75% are never auto-published, regardless of star rating.
• AddedSpam detection flags promotional reviews and drafts a report-to-Google response instead.
• ImprovedBrand voice preview updates live as you move the tone sliders.

• AddedProfile health scoring with weighted checks and per-check hints.
• AddedWebhook events for review.created and scan.completed.
• ImprovedGrid heatmap points are now keyboard-navigable and screen-reader labelled.

• AddedCompetitor set with side-by-side rating, review volume and average rank.
• ImprovedScans are calibrated so the keyword table and the heatmap can no longer disagree.
• FixedLanguage preference was lost on hard refresh.

First public release

• AddedReview inbox with sentiment and intent analysis.
• AddedBrand-voice reply generation with keyword scoring.
• AddedGeo-grid ranking scans up to 13×13.
• AddedEnglish and Croatian throughout.