The Gumshoe Playbook breaks AI visibility work into five steps: establish your baseline, read your results, find your gaps, take action, and see what's working. Each step has its own KPIs worth watching, pulled from your project (a Visibility Audit and the four pillar audits underneath it). Below is a KPI idea for every step, along with exactly where to find it. For the case on why tracking these over time matters, see our article on how a Gumshoe subscription proves ROI to your whole company.
What should I track to establish my baseline?
- Overall brand visibility: your starting share of AI conversations where your brand is mentioned at all. Find it: Visibility Audit dashboard.
- Persona, topic, and model visibility: your starting point broken out by audience, subject area, and AI model. Find it: Visibility Audit dashboard.
- Competitive Leaderboard visibility: where you rank against competitors from day one. Find it: Competitive Leaderboard.
What should I track to read my results?
- Persona scores side by side: the same brand can show 60 percent visibility for one persona and 5 percent for another, and that gap is where the insight lives. Find it: Visibility Audit dashboard, persona breakdown.
- Competitive Leaderboard mentions and visibility: how often you and each competitor come up, compared directly. Find it: Competitive Leaderboard.
- Raw conversations: the actual prompt and answer behind any score, so you can see exactly what a model said. Find it: Conversations page.
What should I track to find my gaps?
- Persona and topic heatmaps: which persona and topic combinations you're invisible for. Find it: Visibility Audit heatmaps.
- Competitor heatmaps: where a competitor's visibility outpaces yours, by persona or by topic. Find it: Visibility Audit heatmaps.
- Content Audit recommendations: how many topics fall into a retrieval issue, a generation issue, or defend position, and how many recommendations are still open. Find it: Content Audit (formerly known as a report).
- Citation counts by domain: which domains AI cites most for your category, broken out by category like ecommerce, social, and forums, and whether your own domain is one of them. Find it: Citations page or a Citation Audit.
- Technical Audit scores: per-URL crawlability and content retrievability scores, plus schema markup flags. Find it: Technical Audit.
What should I track when I take action?
Using the four pillar audits (Content Audit, Sentiment Audit, Citation Audit, and Technical Audit), you can measure your efforts against the results.
- Content recommendations completed versus total: your progress against the open count. Find it: Content Audit.
- Before and after technical readability scoring: whether a fix actually raised the score for that URL. Find it: Technical Audit.
- Citation count growth: whether your own domain is earning more citations, and whether new domains are showing up. Find it: Citations page or a Citation Audit.
- Sentiment theme counts: whether pro themes are growing and caveats are shrinking after you update or correct content. Find it: Sentiment Audit, Themes tab.
What should I track to see what's working?
- Brand visibility trend: the line that shows whether the numbers are actually moving. Find it: Trends.
- The "% of Results Ranked #1" chart: whether you're winning the top spot more often over time. Find it: Trends.
- Sentiment theme counts over time: run over run, whether the same themes keep showing up or the mix is changing. Find it: Sentiment Audit, Themes tab, compared across runs.
- Citation count over time: whether your domain's citation count is climbing after a content or PR push. Find it: Citations page or a Citation Audit, compared across runs.
Pro tip: KPIs only prove progress when you have more than one data point. This is why you need a project to run regularly so every run adds to the trend line these KPIs depend on.