Build with AI · The Kirkpatrick Model

Audit what you already have

Paste in what you already have, and get a prompt that reviews it against the Kirkpatrick Model and suggests fixes.

TL;DR

The Kirkpatrick Model breaks training evaluation into four levels: Reaction (did they like it), Learning (did they learn it), Behavior (are they using it at work), and Results (did it move a business number). Most teams only measure the first two, which can’t prove training worked. The New World version flips the order: start with the result you want, then work backward to the behaviors, skills and content that get you there.

Take the full course

What could go wrong

Double check the details. Make sure you’re telling the AI the right product, feature, or audience. A prompt built on the wrong specifics will confidently give you the wrong output.
Check your spelling and terms. Typos in product names, tier names, or feature names will get picked up literally. Read your inputs before you submit.
Be specific, not vague. The more detail you give about your business, your customers, and your goal, the better the output. Vague inputs get vague prompts back.
Use voice dictation to move faster. Tools like Wispr Flow (or your phone or computer’s built-in dictation) let you talk through your inputs instead of typing, which is especially useful for the longer fields.
This gets you a strong starting point, not a finished product. Review what the AI gives you before you use it with a real team or customer.

← All Build with AI options