Skip to main content

1. Custom analytics and graphs

Prompt: “Using Octolens, calculate the share of voice for BRAND vs. COMPETITOR1, and COMPETITOR2, over the last 90 days. Return a share of voice graph (pie chart) with brand, total mentions, and percentage share of total mentions.” What this does:
  • Checks all your brand and competitor mentions from the last 90 days
  • Creates a custom graph
Recommendation: for heavier analytics searches and graph creation go for a strong model like Claude Opus 4.6 or GPT 5.3
Share of voice graph example

2. Weekly Reputation Monitoring

Prompt: “Show me all mentions from the last 7 days, grouped by sentiment. Highlight any negative mentions that need immediate response.” What this does:
  • Pulls mentions from your tracked keywords
  • Uses AI to analyze sentiment
  • Surfaces urgent issues
Best for: Weekly stand-ups, team reviews

3. Competitive Intelligence

Prompt: “Compare how people talk about [Your Brand] vs [Competitor] on Reddit this month. What are the key differentiators?” What this does:
  • Analyzes language patterns
  • Identifies positioning gaps
  • Surfaces competitive advantages
Best for: Product marketing, positioning strategy

4. Product Feedback Analysis

Prompt: “Extract all feature requests mentioned on any platform in the last month. Group by frequency and urgency.” What this does:
  • Consolidates feedback across platforms
  • Prioritizes by volume
  • Identifies patterns
Best for: Product teams, roadmap planning

5. Support Issue Detection

Prompt: “Alert me to any mentions containing words like ‘bug’, ‘error’, ‘not working’, or ‘broken’ in the last 24 hours.” What this does:
  • Real-time issue detection
  • Proactive support outreach
  • Reduces escalation risk
Best for: Support teams, DevOps

6. Campaign performance

Prompt: “Show me all mentions during our product launch week (Feb 1-7). How did sentiment change day by day?” What this does:
  • Tracks launch momentum
  • Measures campaign impact
  • Identifies amplification opportunities
Best for: Marketing teams, launch retrospectives

Prompting Tips

Be specific about time ranges:
  • “Last 7 days” is better than “recently”
  • “January 2026” is better than “this month”
Specify platforms when relevant:
  • “Reddit and Hacker News” vs. “all platforms”
  • Helps AI focus on the most relevant data
Ask for structured outputs:
  • “Group by sentiment”
  • “Summarize in bullet points”