ChatGPT

A growing share of commercial questions is now answered as prose that names two or three brands and nobody else, and most companies have no idea whether they are one of them. This guide replaces the guesswork with a repeatable method: a frozen 40-prompt set across four categories, a 0-to-3 scoring rubric, and five metrics — presence rate, share of voice, citation rate, answer position and framing. It covers how ChatGPT, Gemini and Perplexity differ, what your server logs can and cannot tell you, the levers that actually move the numbers, and a 90-day plan to get from a baseline to a trend.