Brand, Competitors & Sentiment

Fix Ghost Competitors in AI Visibility Manager

🔍 What Is a Ghost Competitor? A ghost competitor is a domain that appears in your AI Visibility summary or reports but is missing from the competitor management area — meaning you cannot select or delete it through the normal interface. This is a known bug that can affect any tracked site. For example, a competitor like hardscaping.com may show up in the AI Visibility summary for your client's site, yet be completely absent when you open the competitor management area to remove it. 🧭 How to Navigate to AI Visibility Competitor Management 1. Log in to Search Atlas and select the correct client project. 2. In the left sidebar, click AI Visibility (URL: /llm-visibility). 3. Open the competitor management area within the AI Visibility dashboard. 4. Review the list of competitors displayed. If a domain shown in the summary is missing here, you are experiencing the ghost competitor issue. ⚠️ What Is Happening This is a platform-side bug in which a competitor domain becomes visible in the AI Visibility summary but is blocked from appearing in the competitor management area. As a result, the domain cannot be removed through the standard interface. Our team needs to investigate and resolve this on the backend. 🛠️ How to Escalate This Issue Because this is a backend bug, it cannot be resolved through self-serve steps. To get it fixed as quickly as possible, please have the following information ready before contacting support: - The exact domain name of the ghost competitor (e.g. hardscaping.com). - The name of the affected client project in Search Atlas. - The date and approximate time you first noticed the competitor appearing in the summary. - A screenshot showing the domain in the AI Visibility summary and its absence from the competitor management area, if possible. ✅ What to Expect Once you contact support with the details above, our team will investigate the data mismatch on the backend and remove the ghost competitor from your account. You will be notified when the issue has been resolved and the competitor no longer appears in your AI Visibility summary. If you need further assistance, open the chat widget in the bottom-right corner of the platform and type human teammate to be connected with a member of our team.

📊 AI Visibility Sentiment Tab: Occurrences Explained

📌 Overview The AI Visibility report in Search Atlas tracks how your brand appears across AI-generated responses. Inside the Sentiment tab, you will find several metrics that together describe the tone and frequency of your brand mentions. One metric that customers frequently ask about is Occurrences. This article explains exactly what it means and how to use it. 🔍 What Does 'Occurrences' Mean? Occurrences is the total count of times a specific sentiment category — Positive, Neutral, or Negative — was recorded for your brand across all tracked AI responses within the selected date range and query set. In simple terms: every time an AI model responded to a tracked prompt and that response contained a mention of your brand with a detectable sentiment, that counts as one occurrence. The Sentiment tab groups those counts by sentiment type so you can see how often your brand is being described positively, neutrally, or negatively. 📐 How Occurrences Are Calculated Search Atlas processes each AI-generated response that mentions your brand and runs sentiment analysis on the surrounding context. Each analysed mention produces exactly one sentiment label. The platform then tallies those labels: - Positive occurrences — mentions where the surrounding context reflects a favourable tone toward your brand. - Neutral occurrences — mentions that are factual or descriptive with no clear positive or negative tone. - Negative occurrences — mentions where the surrounding context reflects an unfavourable tone toward your brand. The total of all three sentiment categories combined equals the total number of brand mentions analysed in that reporting period. If you see 80 Positive, 15 Neutral, and 5 Negative occurrences, that means 100 brand mentions were processed and scored. 🔗 How Occurrences Relate to Sentiment Score The Sentiment Score shown in the same tab is a weighted percentage derived from your occurrences data. It reflects the proportion of positive mentions relative to all analysed mentions. A higher share of Positive occurrences drives a higher Sentiment Score. Occurrences are the raw counts behind that score. Monitoring occurrences over time helps you understand whether a shift in your Sentiment Score is driven by a rise in negative mentions, a drop in positive ones, or a change in overall mention volume. 📋 How to View Occurrences in the Sentiment Tab 1. In the left sidebar, navigate to AI Visibility. 2. Open the report for the brand or project you want to review. 3. Click the Sentiment tab at the top of the report. 4. The Occurrences column displays the count for each sentiment category — Positive, Neutral, and Negative. 5. Use the date range selector to compare occurrences across different time periods. 💡 Tips for Using Occurrences Effectively - Track trends over time — a sudden spike in Negative occurrences may signal a new AI model is describing your brand unfavourably in response to specific queries. - Compare competitors — if competitor tracking is enabled, you can benchmark their occurrence distribution against yours to understand relative sentiment positioning. - Cross-reference with Mentions — the Mentions metric counts how many AI responses included your brand at all, while Occurrences counts how many of those mentions carried a scoreable sentiment signal. A large gap between the two indicates some mentions lacked enough context for sentiment classification. - Filter by query or topic — narrowing the report to a specific query cluster helps identify which topics are generating negative sentiment, so you can adjust your content or brand messaging accordingly. 📖 Related Metrics in AI Visibility Reports To get the full picture of your AI Visibility performance, use Occurrences alongside these other metrics available in the report: - Rank — the average position at which your brand appears in AI-generated responses. - Citations — the number of times your domain is cited as a source in AI responses. - Share of Voice — the percentage of AI responses in your tracked query set that mention your brand compared to all brands mentioned. - Mentions — the raw count of AI responses that included at least one reference to your brand. - Sentiment Score — the weighted positive sentiment percentage calculated from your Occurrences data. 🆘 Need Help? If you need further assistance, open the chat widget in the bottom-right corner of the platform and type human teammate to be connected with a member of our team.

🔄 Rescan AI Visibility Competitors

🎯 When to rescan competitors After manually adding competitors in AI Visibility, you can start a manual scan to include them in your visibility results. You do not need a separate top-right Rescan button. 🚀 Start a manual rescan 1. Open the AI Visibility page in Search Atlas. 2. Confirm that your manually added competitors appear in the competitor list. 3. Use the manual scan or rescan control available directly on the AI Visibility page. 4. Confirm the action if prompted. ⏱️ Daily scan limit The manual AI Visibility scan can be triggered once per day. If you have already started a manual scan today, wait until the next day before starting another one. Newly added competitors are included when the next scan runs. Results may take some time to update after the scan begins. 💬 Need help? If you need further assistance, open the chat widget in the bottom-right corner of the platform and type human teammate to be connected with a member of our team.

✅ Share-of-Voice & Sentiment Dashboard Data Gaps — May 12–18, 2026 (Resolved)

The automated backfill for May 12–18, 2026 snapshot gaps has been deployed. Your Share-of-Voice and sentiment dashboards are restoring missing data points in stages — no action is required on your end. 🔍 What Happened: May 12–18, 2026 Between May 12–18, 2026, provider downtime and task queue stalls disrupted daily snapshot collection for user-tracked projects. As a result: - Share-of-Voice trend lines showed empty or broken segments for the affected dates. - Sentiment trend charts rendered no data points for those days. - Historical comparisons spanning May 12–18 appeared incomplete or inaccurate. Who is affected: Any account with user-tracked projects that relies on Share-of-Voice or sentiment dashboards for the May 12–18, 2026 date range. All underlying project data is intact — only the snapshot collection step was disrupted. 📊 Additional bar-chart display fixes (separate from snapshot gap) As part of the same sprint, two customer-visible bar-chart display bugs were also identified and resolved: - Sentiment by Platform (LLMV-495): Was showing only the most-recent snapshot instead of the full selected date range. - Visibility by Platform (LLMV-479): Same display bug — only the most-recent snapshot was shown instead of the full selected date range. These bugs are unrelated to the snapshot collection outage but were fixed in the same sprint. If your Sentiment by Platform or Visibility by Platform bar charts were showing incorrect data outside the May 12–18 window, that was this separate display bug — now also resolved. 🛠️ No Action Required The backfill is restoring missing data points in stages across all accounts. Most accounts will see full restoration within 24 hours; accounts processed later in the queue may take longer. Missing snapshots cannot be re-triggered or manually filled from within the Search Atlas interface — recovery is fully automated. If your dashboards still show gaps after 24 hours, contact support via the chat widget in your Search Atlas dashboard. 📡 Engineering Status - Root cause confirmed: Snapshot collection failed between May 12–18, 2026, due to provider downtime and task queue stalls. - Fix developed: An automated gap-detection and backfill system was built to identify and restore all missing snapshots across affected accounts. - Deployed — LLMV-460: The backfill solution has been released and is actively restoring affected data points in staged batches. - Secondary bug patched: A secondary bug (backfill skipping days with in-flight failed snapshots) was identified and patched. If gaps persist beyond 24 hours, this edge case may apply — contact support so engineering can trigger a targeted re-sweep for your account. - No permanent data loss: All underlying project data is fully intact. - Status as of May 22, 2026: Fully resolved and deployed. 📞 When to Contact Support A member of our support team will proactively reach out to confirm the restoration status for your specific projects. Contact support if: - Your Share-of-Voice or sentiment dashboards still show missing data 24 hours after this update. - You notice data gaps outside the May 12–18, 2026 window. Reach support via the chat widget in your Search Atlas dashboard or through your account contact directly. 💬 Resolved Support Conversations The following conversations were escalated to a human support agent during this incident and resolved directly. Conversation 1 — Missing Share-of-Voice Data (Urgent Client Report) User: Hi, my Share-of-Voice charts are completely blank for the last week. I need this data for a client presentation today — is this a known issue? Support (automated): We are aware of a data collection issue impacting Share-of-Voice and sentiment dashboards for May 12–18, 2026. Our engineering team is actively working on a resolution. User: This isn't good enough — I need someone to actually look at my account. This is urgent. — Escalated to human agent — Support Agent (Alex): Hi, thanks for your patience. I've reviewed your account and confirmed your tracked projects are affected by the May 12–18 snapshot gap. I'm flagging your account for priority processing in the backfill queue — you should see the missing data restored within 2–4 hours. I'll follow up with you directly once it's confirmed on our end. User: Thank you, I really appreciate that. Support Agent (Alex): Quick update — all your Share-of-Voice data for May 12–18 has been restored. Your trend lines should now display complete data. Please refresh your dashboard and let me know if anything still looks off. User: Just checked — everything looks perfect now. Thank you! Resolution: Data restored via priority backfill. Confirmed resolved by agent. Conversation 2 — Sentiment Dashboard Showing No Data Points User: My sentiment trend chart has no data at all for the past several days. I've tried refreshing and clearing my cache but nothing changes. Support (automated): This is likely related to a known snapshot collection issue affecting May 12–18, 2026. No action is needed from your end — a fix is being deployed. User: I've seen that message before and the data still isn't there. Can a real person check this? — Escalated to human agent — Support Agent (Maria): Hi! I've reviewed your account. Your sentiment dashboard is affected by the May 12–18 data gap and your projects are included in the active backfill. Restoration is in progress — your data should be fully populated within the next few hours. I've added a note to monitor your account specifically and will reach out if anything doesn't resolve automatically. User: Okay, thank you for actually looking into it. Support Agent (Maria): Following up — the backfill has completed for your account. All sentiment data for May 12–18 is now restored and your trend charts should be accurate and complete. Let us know if you have any other questions! User: Great, all good now. Thanks! Resolution: Sentiment data restored via automated backfill. Agent confirmed completion directly with user. 🎯 The backfill for May 12–18 snapshot gaps has been deployed — your Share-of-Voice and sentiment dashboards are restoring complete, accurate trend data in stages with no steps required on your end. Most accounts will see full restoration within 24 hours; if gaps persist after that, contact our support team directly.