You can't tell if anything is working unless you ask the same questions the same way over time. So this week I built the boring but important part: a monthly prompt log for Cedar Park. What it is: a simple sheet with the same local questions, asked every month in Google's AI Overviews, ChatGPT, Gemini, Perplexity and the map pack (the 3 businesses Google shows with a map). The questions (first draft): 1. Best roofer in Cedar Park TX 2. Cedar Park dentist accepting new patients 3. AC repair Cedar Park TX 4. Who does local SEO in Cedar Park? What I log for each: date, tool, logged in or out, location, who got named, whether they're actually in Cedar Park, and a screenshot. Why it matters for you: you can do the same for your business. Pick 3 questions your customers ask, run them on the 1st of each month, and screenshot them. In 3 months you'll see patterns that one search never shows. What I learned from the first run: on Oct 29 I ran the roofer question for Monday's teardown. Only one roofer showed up everywhere, Google and ChatGPT named different lists, and a more specific question (hail damage) pulled in Austin companies. That's exactly why one search is a snapshot, not a verdict. What I don't know yet: how much answers drift month to month with no changes at all. That's the first thing this log should tell us. I'll share the first comparison in early December. I'm being upfront that question 4 is about my own work. If AlexSEO isn't named, I'll say so and show what I change. Below: the empty log template.