Understanding the dashboard
A guided tour of every widget on your home dashboard.
Last updated: 2026-05-05
TL;DR — The dashboard is your single-pane view of how your brand performs across AI engines. Top widgets show Visibility, Share of Voice, Top Brands, and Co-occurrence. Three global filters — Country, Date range (default 30 days), and Simulation — change every widget at once. The dashboard prefers Unaided prompts for the most organic, unbiased view.
Why the dashboard exists
Every AI Scan you run produces hundreds or thousands of data points: brand mentions, ranks, sources, sentiment, topic context. Raw data isn't insight — you need aggregation, segmentation, and time-series. The dashboard takes everything from your scans (filtered by your active brand, country, date range, and selected scans) and presents the 8–10 metrics that actually matter for AI search visibility, in the order you'd want to see them in a weekly review.
If you only have 60 seconds, the AI Score card at the top tells you whether you're winning or losing. Everything below explains why.
The widgets, top to bottom
AI Score
Your overall AI Search visibility — a 0–100 score blending rank quality, how often you appear within a scan, and rank stability, scaled by Coverage across all selected models in the chosen window. Coverage and Mention Rate sit alongside it. See How Pranas measures AI visibility for the formula. The sparkline shows trend; arrows indicate week-over-week change.
Share of Voice (SOV)
The percentage of brand mentions across all responses that go to your brand vs. competitors. SOV = (your mentions) / (your mentions + competitor mentions). See Share of Voice.
Top Brands
Ranked list of the most-mentioned brands across your queries. Toggle Competitors only to filter to brands you've explicitly registered as competitors — this gives a much cleaner picture once your competitor list is set up. See Top Brands.
Co-occurrence
Which brands are mentioned alongside yours in the same response. High co-occurrence with a competitor means you're often considered together — a positioning signal you can leverage. See Co-occurrence.
Topic Visibility
How well you appear across the topic clusters relevant to your category. Identifies topics where you dominate vs. topics where you're invisible.
Narrative Drivers
The themes and attributes models associate with your brand vs. competitors. Useful for spotting positioning gaps and brand-perception drift.
Sources Summary
Top citation sources (URLs and domains) that AI engines reference when discussing your category. See Sources overview.
Brand Perception Summary
A snapshot of sentiment, perception drivers, and recommended actions. Click into Brand Perception for the full analysis.
Action Center Summary
The 3–5 highest-priority recommendations Pranas has generated based on your latest scans. Each is one click away from a content brief or remediation plan in the Action Center.
The three global filters
Every widget on the dashboard responds to the same three filters at the top of the page:
| Filter | Default | What it does |
|---|---|---|
| Country | "All countries" | Scopes data to one country, or "All". Country is set per-scan; this filter narrows what you see. |
| Date range | Last 30 days | Time window for all metrics. Common choices: 7d, 30d, 90d, all-time. |
| Simulation (AI Scan) | All scans | Limit to a single scan or a hand-picked set. Defaults to the last 30 days of scans. |
Important behavior: filters are persistent across page navigation within the dashboard area. Switching to Brand Perception or Sources also respects the same filters where applicable.
Why "Unaided" matters
Pranas tags every prompt as one of: Unaided, Aided, Comparison, or Branded. The dashboard prefers Unaided prompts (queries where the user does not name your brand, e.g. "best customer feedback tools") because they reflect how prospects actually discover you — without prior knowledge.
- Aided prompts ("is Acme Feedback any good?") inflate visibility because the model is forced to discuss you.
- Comparison prompts give useful relative-positioning data but skew SOV.
- Unaided prompts give the truest signal of organic AI visibility.
Some widgets (Top Brands, Co-occurrence) are computed exclusively from Unaided. Others show a blended view but flag the breakdown. See Prompt types.
Worked example
It's Monday morning. Maya opens the dashboard:
- Filter setup: Country = US, Date = last 30 days, Scans = all.
- AI Score: 47 (↑ 6 vs. previous 30 days). Good — the new content she shipped is moving the needle.
- SOV: 12% (↑ 2 pts). Acme is gaining ground on Canny.
- Top Brands (competitors only): Canny 38%, ProductBoard 24%, Acme 12%, Frill 8%. She's #3 — clear path to #2.
- Co-occurrence: Acme appears with Canny in 41% of responses. Strong "considered together" signal — she should publish a Canny vs. Acme comparison page.
- Action Center: 4 recommendations, top one is "Publish a Canny vs Acme comparison page".
In 90 seconds she has a weekly view, a thesis, and a next action.
Edge cases & gotchas
- First-time empty state: until your first scan completes, widgets show empty/skeleton states with a CTA to run a scan.
- Country filter shows "no data": that country had no scans in the selected window. Check your scan history or run a scan scoped to that country.
- Numbers don't match between dashboard and Results page: the dashboard applies filters; Results shows raw scan data. Always check filters first.
- Top Brands shows non-competitors: toggle Competitors only on. AI engines mention many tangentially-related brands; this filter focuses on the ones you actually compete with.
- Sparklines look noisy: short windows (7d) on small scan volumes will be jumpy. Widen the window or run more scans for stable trends.
Troubleshooting
| Symptom | Likely cause | Fix |
|---|---|---|
| All widgets empty | No scans completed yet, or filters too narrow | Run a scan, or widen date range |
| AI Score = 0 but you saw mentions in Results | Filter mismatch (country, scan, date) | Reset filters |
| Top Brands missing your brand | You appear in <1% of responses, or only in Aided/Branded prompts | Check Unaided prompt coverage |
| Widget shows old data | Cache; happens rarely after scan completion | Hard refresh (Ctrl/Cmd+Shift+R) |
Related articles
Still need help?
Email contact@pranas.co — a screenshot of your dashboard and the filter state helps us help you faster.