About this module
Learners get a usable mental model in this section. Learners see how common prompting mistakes that waste time affects a real workplace task. They practice spotting vague requests, missing audience, unclear source material, broad scope, and no review criteria and compare the result with the risks that come from accepting generic output because the task was too loosely framed. The module ends with a simple standard: know the purpose, check the output, and fix prompt problems before they create extra editing work.
Key takeaways
You already use A.I. every day, but small habits in how you ask are silently capping the quality of what comes back. Let's break them. Most bad answers do not come from a weak A.I. model, they come from a weak prompt.
Let's fix that, one mistake at a time. First up, vagueness. When you leave out what you're trying to accomplish, the model has to fill in every blank on its own, and it usually guesses wrong.
Here's what that looks like in practice. Write something about marketing, is a prompt with no topic, no audience, and no format, so the output can't have one either.
Second, dropping the context the model actually needs, like your audience, your budget, or what you've already tried, so it defaults to generic, one-size-fits-all advice.
How do I get more customers, sounds reasonable, but with no industry, no budget, and no channel mentioned, the advice comes back just as generic as the question.
Third, cramming several unrelated jobs into a single message. The model spreads its attention thin across all of them, and none come back fully developed. Four separate jobs crammed into one prompt means none of them get the depth they need. Break big asks into smaller, sequenced requests instead.
Fourth, staying silent on how you want the answer shaped. Say nothing, and you'll get whatever default structure the model reaches for, which rarely fits your actual need. Summarize this report gives no hint about format, so you get one dense paragraph instead of the five bullet points you actually needed for the meeting.
And fifth, stopping after one try. The model can revise, tighten, or completely rework an answer in seconds, but only if you actually ask it to. The biggest missed opportunity isn't a bad first prompt, it's never sending a second one. A single follow-up like, make it shorter, often does more than a perfect first try.
Every one of these mistakes has the same fix: add a goal, an audience, a length, and a format. Specific prompts consistently beat vague ones. To recap: vague prompts, missing context, overloaded requests, missing format, and skipping iteration are the five mistakes to watch for. Catch these five mistakes and your prompts improve almost overnight.
Next up, advanced prompting strategies for when good prompts aren't enough.
