For complex tasks, ask AI to outline the steps before doing anything.
Same complex request. One jumps straight in. The other maps out the work first.
AI: First, let's export your WordPress content. Go to Tools > Export and download the XML file...
[Continues with detailed instructions, mixing setup, content migration, and deployment steps without clear structure]
Already doing things. But what's the full picture?
AI: Here's the migration plan:
Ready to start with step 1?
Now you see the full scope before committing.
When AI jumps straight into execution, it can miss steps, go down the wrong path, or solve a problem you didn't actually have. By the time you realize something's off, you've already invested time following instructions that won't lead where you need to go.
Asking for a plan first creates a checkpoint. You see the full scope of work before anything happens. You can catch misunderstandings, add missing steps, or redirect the approach — all before any real work begins.
For complex tasks, ask AI to show you the plan before executing anything. Review it, adjust it, then work through each step together.
Once you have a plan you're happy with, work through it step by step. This pairs well with One Thing at a Time — focus on each step fully before moving to the next.
As you go, the plan might change. That's fine. The point isn't rigid adherence — it's having a map so you know where you are and where you're heading.