My name's Brandon Galey, and this is the Claude Genius podcast, where I teach you the top growth hacks from Claude Power users in short, no-fluff episodes.

Every episode has a companion substack post that goes deeper with copy-and-paste prompts, meta-promps, or skills.

That means you don't need to take any notes

because everything is in the companion post

that I link to in the episode description.

Now let's get into this episode

where I'll be teaching you how to 10x your output

when you toggle on research and clod.

Here's a number that should bother you.

When you flip on research and Claude, the single run burns about 15 times the tokens of a normal chat.

15 times. And most people get back something that's maybe twice as useful as a regular question in chat would have given them.

So you're paying a fortune and pocketing the change.

Here's the part that nobody believes.

That gap is not the model's fault.

It's ours.

The majority of the time the reason research disappoint, comes down to how he asked it. And the good news hiding inside that bad news is that the entire

fix comes down to nine things you tell Claude to do before it runs the research. All nine of these

fall into three buckets.

The first bucket is vague scope.

You type Research X with no number, no time frame,

no rules for what counts as a good source.

Claude has no clue whether you want five examples or 50,

...