Skip to main content
Visibility Analytics is the core monitoring dashboard in Lightsage. It runs your defined prompts across 17 AI platforms on a daily or alternate-day schedule and aggregates the results into a single view: how often AI mentions your brand, where you rank relative to competitors, and which prompts are driving or losing visibility.

​
What Visibility Analytics tracks

Every run captures three top-level metrics across all active platforms:
  • Visibility Score — a composite rate measuring how frequently your brand appears in AI-generated responses for your tracked prompts
  • Share of Voice — your mention count as a percentage of all brand mentions (yours plus competitors) for the same prompt set
  • Average Position — where your brand appears in a response, averaged across all platforms that mentioned you
These roll up into a single dashboard so you can spot trends without switching between platforms.

​
The two types of AI platforms

Lightsage tracks 17 platforms split across two categories. Understanding the difference matters because they drive different business outcomes.

Answer engines (7)

ChatGPT, Google AI Overviews, Google AI Mode, Gemini Search, Perplexity, Claude, Microsoft CopilotThese platforms answer questions in prose. Visibility here drives brand awareness — developers learning your name before they start building.

Coding agents (10)

Claude Code, OpenAI Codex, Cursor, GitHub Copilot, Gemini CLI, OpenClaw, OpenCode, Hermes, Pi, KiloThese platforms write and execute code. Visibility here drives adoption — agents recommending and implementing your API directly in a developer’s project.
For API and developer tool companies, coding agent visibility is where adoption actually happens. Answer engine visibility builds awareness; agent visibility converts. Track both, but weigh agent metrics heavily when prioritizing improvements.

​
Setting up prompt tracking

Prompts are the queries Lightsage sends to each AI platform to check whether your brand is mentioned. The quality of your prompts directly determines the quality of your data.

​
The four prompt types

Prompts about your company or product by name.Examples:
  • “What is Acme API?”
  • “How does Acme API work?”
  • “Is Acme API reliable?”
Purpose: Establish a baseline mention rate. If you score low on brand queries, the AI platforms either don’t know about you or have outdated information.

​
Adding prompts

1

Go to Visibility Analytics

Open the Visibility Analytics section in your Lightsage dashboard and select Manage Prompts.
2

Add brand queries first

Start with 3–5 direct brand queries. These establish your baseline and are the fastest to set up.
3

Add category and comparison queries

Add 5–10 category queries covering your main use cases. Then add at least one comparison query per major competitor.
4

Add implementation queries for coding agents

Add 3–5 implementation prompts — one per top use case. Phrase these as a developer would actually type them into a coding agent.
5

Select platforms and schedule

Choose which platforms to run each prompt on. Answer engine prompts cost 1 credit per run; coding agent prompts cost 5 credits per run. Set the frequency to daily for high-priority prompts or alternate days for lower-priority ones.
Write implementation queries in first person and specify a framework: “Add Acme payments to my Express app” performs better than “How to integrate Acme API.” Coding agents respond to the same phrasing developers use when they’re actively building.

​
Understanding the Visibility Score

Your Visibility Score is a number from 0 to 100 that represents your weighted brand mention rate across all tracked platforms and prompts. The score accounts for:
  • Mention presence — whether your brand appears at all in a response
  • Position — responses where you’re mentioned first score higher than ones where you’re buried at the end
  • Platform weight — coding agent mentions weight more heavily than answer engine mentions for developer tool companies
  • Prompt relevance — implementation and comparison prompts weight more than brand queries
A score of 60 doesn’t mean you’re mentioned in 60% of responses. It’s a composite index. Use it to track direction over time rather than interpret as an absolute percentage.
Scores vary significantly by category maturity and competition density. As a rough guide:
  • 70–100: Strong visibility. You’re consistently recommended across most relevant prompts.
  • 50–69: Moderate visibility. You appear in many prompts but lose significant share to competitors.
  • 30–49: Low visibility. Agents know you exist but rarely recommend you unprompted.
  • Below 30: Minimal visibility. Your API likely has discoverability issues (robots.txt blocking, no llms.txt, thin documentation).
Score drops have a few common causes: a competitor improved their visibility (check Share of Voice), your documentation changed in a way that reduced AI crawlability, or an AI platform updated its model and your training data presence decreased. Check the per-platform breakdown to isolate which platform drove the drop.
Your Visibility Score recalculates after each prompt run. If you run prompts daily, your score updates daily. If you run on alternate days, the score updates every two days.

​
Share of Voice explained

Share of Voice (SoV) measures what percentage of total brand mentions in a prompt set belong to you versus your tracked competitors. Formula: Your mentions ÷ (your mentions + all competitor mentions) × 100 If Claude Code mentions your API 4 times and your two competitors 6 times combined across 10 category prompts, your Share of Voice is 40%.
Share of Voice can be misleading if your competitor set is too narrow. Make sure you’ve added all meaningful competitors in your category — not just the ones you’re actively aware of. Lightsage can suggest competitors based on co-mention patterns in your prompt results.

​
Reading SoV over time

The SoV trend line is more valuable than the current number. Look for:
  • Rising SoV: Your improvements are working. Continue the current strategy.
  • Flat SoV with rising Visibility Score: Both you and competitors are growing together. You’re not losing ground, but you’re not winning either.
  • Falling SoV: A competitor is improving faster than you. Open the competitor breakdown to see which prompts they’re winning.

​
Prompt run frequency

Daily runs give you the tightest feedback loop when you’re actively making changes. Alternate-day runs are more credit-efficient for stable prompts where you’ve already established a baseline.

​
Metrics reference

​
Tips for choosing effective prompts

Match real developer intent

Write prompts the way a developer would actually type them — not how you’d describe your product in marketing copy. “Best auth API” outperforms “What authentication solutions are available?”

Cover multiple frameworks

Add implementation queries for each major framework your users work in: Next.js, Express, FastAPI, Rails. Agent recommendations are framework-specific.

Include your competitor names

Comparison queries like “[Your API] vs [Competitor]” produce some of the most actionable data. They directly reveal whether AI systems recommend you in head-to-head comparisons.

Avoid overly broad queries

“Best API” produces noisy results across too many categories. “Best payment API for SaaS” or “Best auth API for Node.js apps” gives you targeted data on prompts you can actually win.

​
Reading and acting on results

After each run, the Visibility Analytics dashboard shows a breakdown by prompt, platform, and time period. Use this workflow to turn data into action:
  1. Check the trend line first. Is your Visibility Score up, down, or flat over the past 30 days?
  2. Look at Share of Voice by prompt type. Are you losing on category queries, comparison queries, or implementation queries?
  3. Identify which platforms are dragging you down. Low scores on coding agents but high on answer engines suggests a discoverability problem (robots.txt, llms.txt). Low scores everywhere suggests a content problem.
  4. Open the Opportunities Engine. Lightsage automatically surfaces prioritized fixes based on your Visibility Analytics data. See Opportunities Engine for details.