Prompt engineering is arguably the most important skill in this era of AI tools.
There are so many elements to it though and it can be hard to know which ones are more important than others.
From my own experience and other engineers’ documentation, feedback, videos, etc. I will say that if I had to pick one thing that matters more than everything else it would be adding examples to your prompts.
Think about it, the AI tool is basically trying to figure out what you are looking for. All of the elements that you add to your prompts (i.e., context, personas, tone, audience, etc.) are directing the model towards what you want to see. If you just tell it what you want to see, by giving an example, it doesn’t have to guess. It now just has to match the new information with the example(s).
Multiple examples are better than one, but even one can help a lot.
And I realize that it takes longer to add an example to your prompt, but you’ll have to decide if it’s worth spending the time upfront. The alternative might be spending more time with follow-up prompts, or getting frustrated with the output that you get, and walking away.
TL;DR - add examples to your prompts to improve the quality of your output