6 Techniques for Effective Prompting
About this resource
The gap between a mediocre AI answer and a genuinely useful one is almost never the model. It is the instruction. This free one-page handout shows six prompting techniques as matched before-and-after pairs, so the difference is visible rather than described. You see the vague request, then the specific one, and the reason the second works.
The techniques cover the things that reliably change output quality: giving the model a role, supplying context it could not infer, being explicit about format and length, including an example of what you want, breaking a complex request into steps, and asking it to check its own work. It is drawn on Anthropic’s published prompt engineering guidance and reduced to a single page you can pin next to a monitor. Nothing here takes longer to apply than the time it saves on the first attempt.
Every technique is shown twice, once badly and once well. Reading the pair takes about fifteen seconds and it transfers faster than any explanation of the principle behind it.
How to use it
Print it and put it where the work happens. Prompting improves through repetition against a visible reference, not through a single training session. Pick one technique, apply it to a task you already have this week, and compare the result against how you would have asked before. The pair format exists so you can check your own instruction against a known-bad version and see which one it resembles.
Inside the download
Six before-and-after pairs make the change concrete instead of theoretical.
Why "you are an experienced auditor" changes the answer, and when it does not.
The information the model cannot infer and will invent if you leave it out.
How to stop getting an essay when you needed three bullet points.
Common questions
Specificity in four areas: who the model should act as, what context it needs that it cannot infer, what format and length you want back, and an example of what a good answer looks like. Vague requests produce vague answers, and the model will fill any gap you leave with something plausible rather than something correct.
Add context and an example. Most disappointing answers come from instructions that assumed the model knew something about your situation, your audience, or your desired format. Including two or three sentences of context and one example of the output you want changes results more than any other single adjustment.
Yes, as a practical layer rather than a foundation. Technique produces measurably better results and it is quick to learn. It is worth keeping in perspective, though, because specific techniques change with each model generation while judgment about when to use these tools and what to verify does not.
Free. No email list, no follow-up sequence.
More from the resource library, chosen because they pair well with this one.
Two practical mnemonics that anchor everyday AI use: better prompts in, verified answers out.
Customizable prompt templates for customer service, administration, healthcare, retail, finance, and field operations.
What these tools do genuinely well and where they reliably fail, so you can assign work correctly the first time.
Tell us a little about yourself and this document is yours. We don't have an email list. We collect this so we know who's finding our resources useful, and that's it.
You're downloading
this document
Your information is kept confidential and never shared with anyone.
