> ## Documentation Index
> Fetch the complete documentation index at: https://docs.mentioned.com.au/llms.txt
> Use this file to discover all available pages before exploring further.
# Understanding Mentioned Scan Results: Keywords and Sentiment
> Learn how to read a completed Mentioned scan, from the summary card through to keyword-level AI responses, sentiment, and competitor overlays.
Each time Mentioned runs a scan, it queries your tracked keywords across the AI engines included in your plan and records every response. The full results of that scan are stored and accessible at any time from your dashboard. To open a scan, navigate to the scan history list at the bottom of the dashboard and click the row for the scan you want to review. This opens the scan results page, which breaks down exactly what the AI engines said about your brand — and your competitors — for each keyword.
## Scan summary
The scan summary card sits at the top of the results page and gives you a quick snapshot of the scan's scope and overall performance.
Shows the date and time the scan was run, the AI engines that were queried, and the total number of keywords included in that scan.
Displays your brand's AI Share of Voice, Mention Rate, Brand Health Score, and Citation Score as calculated from this scan alone — useful for comparing against other individual scans.
The metric scores shown in the scan summary reflect only the data collected during that specific scan, not an aggregated or rolling average. This makes them useful as point-in-time benchmarks, particularly when you want to compare scans before and after a campaign.
## Keyword-level results
Below the summary card, you'll find the keyword results table. Each row represents one tracked keyword and shows at a glance how your brand performed for that query.
The columns in the keyword table are:
* **Keyword** — the exact keyword or phrase that was submitted to each AI engine
* **AI engines** — icon indicators showing which engines mentioned your brand in their response to this keyword
* **Mention type** — whether the mention was **direct** (your brand name explicitly stated) or **indirect** (a reference that clearly implies your brand without naming it)
* **Citation included** — a tick or cross indicating whether your brand was cited as a source or reference in the AI response
### Expanding a keyword row
To read the actual AI-generated text behind a result, click the expand arrow on the left side of any keyword row. This reveals a panel showing the relevant excerpt from each AI engine's response, with your brand's mention highlighted in context.
An indirect mention occurs when an AI engine references your brand in a way that is clearly identifiable — for example, describing your product's unique features, quoting your published content without naming the source, or referring to "the market leader in \[your category]" in a context where your brand is the dominant player. Mentioned's detection logic uses contextual analysis to identify these cases.
## AI engine responses
For each AI engine that was included in the scan, you can view the raw response excerpt it generated for each keyword. This is the actual text the AI returned — not a summary or paraphrase.
Within each response excerpt, Mentioned uses inline highlighting to draw your attention to relevant content:
* **Your brand** is highlighted in blue
* **Competitor brands** are highlighted in orange
* **Citation links** are underlined and marked with a link icon
### Sentiment indicators
Where Mentioned has detected a mention of your brand within a response, a sentiment badge is shown alongside the excerpt:
* 🟢 **Positive** — the mention frames your brand favourably (e.g. recommending your product, citing your expertise)
* ⚪ **Neutral** — the mention is factual or descriptive without a clear positive or negative framing
* 🔴 **Negative** — the mention includes critical language, comparisons that disadvantage your brand, or caveats that may deter users
Sentiment is determined by Mentioned's natural language analysis and is assessed in the context of the surrounding response, not just the sentence containing your brand name.
## Competitor comparison within a scan
If you've added competitors to your Mentioned workspace, you can overlay their results directly onto the scan results view to see how they appeared in the same set of AI responses.
Click the **Compare competitors** toggle near the top of the keyword results table. This is off by default so the view stays uncluttered.
A dropdown will appear listing all competitors in your workspace. Tick the ones you want to include — you can select multiple.
The keyword table will now show mention indicators for each selected competitor alongside your brand, so you can see in a single row which brands appeared for that keyword and on which engines.
Expand any keyword row to see competitor mentions highlighted in orange within the same AI response excerpts, making it easy to see how the AI framed multiple brands in the one response.
## Comparing scans over time
Mentioned's Compare feature lets you select any two scans from your history and view a side-by-side diff of the results, showing what improved, what declined, and what stayed the same.
To compare two scans:
From the scan results page, click the **Compare** button in the top-right corner, or navigate to the scan history list in the dashboard and select **Compare** from the action menu on any scan row.
Choose a baseline scan (the earlier of the two) and a comparison scan (the more recent one). The comparison view will show changes relative to the baseline.
The comparison view highlights increases in green and decreases in red for each metric, keyword, and engine. You can toggle between a summary view (metric-level changes only) and a full keyword-level diff.
Comparing scans from just before and just after a content update, PR campaign, or thought leadership push is one of the most effective ways to measure whether your efforts are improving your AI visibility. Run a scan immediately before you publish, and again two to four weeks later, then use Compare to see the impact.
If your scan results look unexpected — for example, a keyword shows no mentions when you'd expect your brand to appear — there are a few common causes including keyword phrasing, engine coverage settings, or scan configuration. Visit [Running a Scan](/setup/running-a-scan) for troubleshooting guidance and tips on refining your keyword list.