Scans
A scan is the core unit of work in Sightlabs. You define a query, pick your sources, and Sightlabs fetches and analyzes the results. Here is how each part works.
The three scan modes
Every scan starts with a mode selection. The mode controls how Sightlabs queries each source and what it treats as a match.
When to use: Use when tracking a company, product, or app name.
How it works: Searches for exact and near-exact name matches across posts, comments, and AI responses. Catches common misspellings and abbreviations.
Examples: Sightlabs, Notion, Linear
When to use: Use when tracking a person, creator, or executive.
How it works: Searches for the person's name and social handles. Works across platforms that mention people by name, not just @-tag.
Examples: @levelsio, Lenny Rachitsky, Paul Graham
When to use: Use when tracking a topic, trend, or phrase.
How it works: Searches for the exact phrase. Best for narrow, specific phrases. Broad single-word queries return too much noise.
Examples: AI share of voice, no-code tools, B2B SaaS pricing
Source selection
Before running a scan you choose which platforms to include. Selecting fewer sources gives faster results. Selecting more gives broader coverage. Both are valid approaches depending on your goal.
Your plan determines which sources are available. The Starter plan covers social and search sources. Pro and Team add AI platform monitoring (ChatGPT, Perplexity, Gemini, Copilot, Grok). See Sources for the full list.
What happens when you click Run Scan
Parallel fetch
Sightlabs queries all selected sources at the same time. Each source returns a list of matching posts or content snippets.
AI analysis
The full mention set is sent to an LLM pipeline. It assigns per-mention sentiment (positive, neutral, negative), extracts key themes, identifies action items, and flags any PR risks.
Results returned
The scan result appears inline. You see the total count, sentiment breakdown, source distribution, top mentions, and the full AI report.
Cached vs live scans
A small set of popular queries (e.g. Notion, Figma, Linear) return cached results instantly. The result panel shows (cached) next to the scan time when this happens.
All other queries run live against the sources. The result panel shows LIVE when the data was freshly fetched. Scheduled scans (Pro and Team plans) always run live.
Saving and re-running scans
Once you run a scan from the dashboard, it is saved automatically. You can reopen it from your scan history and click Re-run to fetch fresh data with the same query and source settings.
Each re-run adds a new data point to your analytics timeline. This is how Sightlabs builds the 8-week trend view. See Analytics for details.