T: Tune the response

Tuning means refining ChatGPT’s first response with specific feedback about what to keep, change, add, or remove. This happens over a conversation, and is not necessarily a one-time command.

Treat the first response as a draft. Review it for:

Missing information

Unsupported assumptions

Generic recommendations

The wrong level of detail

Weak organization

Wrong tone or format for the task

Then tell ChatGPT precisely what to keep and what to change.

For example:

Phase two feels too ambitious. Limit it to two active projects and allow more time for building relationships with the team. Keep the phase-one goals unchanged.

Other useful follow-ups, depending on what the first response needs:

Identify the weakest part of this response and explain how you would improve it.

Keep the recommendations, but make the reasoning more specific to a B2B SaaS company.

Cite the relevant source for each factual claim drawn from my materials. Flag claims where the supporting source is missing.

Give me a second approach that uses a different underlying strategy.

Rewrite the response using this feedback: [add feedback]

Tuning builds on what the first response gets right. It lets you revise the weak parts without rebuilding the entire response.