Understanding the metrics

1 AI share of voice What share of voice measures in AI answers, and how to read it against mention and recommendation. 2 Mention versus recommendation The difference between being named in an AI answer and being put forward as a choice. 3 The visibility ladder The five-rung outcome scale that every scan result sits on, and what each rung means for your brand. 4 Sources and citations The difference between grounded answers that cite sources and parametric answers that draw on model memory, and why it matters for your AI visibility. 5 Accuracy and fact checks How AI Native checks whether AI answers get your product facts right, what gates the audit, and what the accuracy states mean. 6 Why AI numbers vary between scans AI engines are stochastic, and the same question can produce different answers across runs. Here is why your numbers move, and how to read the signal from the noise. 7 Demo data versus live data What demo mode is for, how it differs from live data, why the two never mix, and when to switch to live. 8 Sentiment in AI answers What sentiment measures in AI answers, how in-AI framing and web-source polarity are separated, and what to do when the number is negative. 9 Demand and value scoring How AI Native grounds each prompt in real search demand and combines it with intent and funnel position to produce a prompt value score. 10 The opportunity score How AI Native combines value, visibility gap, and winnability into a single ranked number that tells you which cell to address first.