What to measure when SEO no longer predicts pipeline

For most of the last decade, a clean relationship held: rank more keywords on page one, get more organic traffic, get more pipeline. That line is now cracked in a lot of categories. Here's what to measure in the world we actually live in.

1. How often AI names you

The headline number: across a tracked set of buyer questions, how often does ChatGPT, Claude, Perplexity, Gemini, or Copilot name your brand? Track weekly. Break down by model. Break down by early-stage (category) vs. late-stage (vendor) questions.

This is the metric that most directly correlates to whether your brand enters the shortlist AI is writing for your buyers.

2. Share of AI answers

Of the answers that name anyone in your category, what percentage include you? This normalizes for popularity of a question and shows whether you're winning head-to-head.

3. Direct, branded, and AI-referred traffic

Organic traffic to your blog is still a signal, but increasingly the interesting line is direct and branded traffic — people who learned about you in AI and then typed your name into a browser. Watch whether that line moves as your AI visibility changes.

4. Pipeline created from AI-influenced opportunities

The measurement that makes executives care: ask every inbound lead during the form or sales process how they heard about you. Track the growing share that says "ChatGPT," "Perplexity," or "someone recommended you — I think I saw you in AI." That line is almost certainly growing faster than you think.

5. Content attribution

Tag every piece of content with the specific questions it's designed to influence. When your mention rate on those questions moves, you know which content moved it. This is how you stop guessing about what's working.

Together, these five measurements give you a real-time picture of whether AI is recommending you. Our clients look at them weekly. Anything on this list that you don't have visibility into is a visibility gap worth closing.