You prompt Claude and wait, guessing whether it understood you, while it works. There is a way to know before the answer arrives.
Claude shows a thinking indicator while it works: "thinking," "building a response," or a short summary of what it's doing. Click it and Claude shows its actual reasoning, not the final answer alone.
Jeff Fargo, founder of Fargo Factor, walks through why the extra few seconds are worth it: reading the reasoning tells you if Claude is on track before the full answer lands. Catch a wrong turn early and redirect, instead of getting a bad answer and starting the prompt over.
It's a habit, not a new tool. Next time you see the thinking indicator light up, click it and read it before you read the answer. Over time your AI-assisted work runs smoother because you're not un-guessing wrong turns after the fact.
Most owners check their AI visibility wrong and think they're fine. fargofactor.ai
