Skip to main content
All CollectionsGenerating text
Top Tips for Text Generation
Top Tips for Text Generation

Optimise your text generation with some of our top tips

Christopher John avatar
Written by Christopher John
Updated over a year ago

We hope through using some of these tips and implementing them as best practice, you will find you are able to extract significantly more value from the text generation tool.

  • Ensure you have a rich and well developed brand library and detailed product descriptions.

  • Add brand assets that the model can draw on to provide content that clearly understands your brand and/or product.

  • Upload documents and URLs such as your company’s marketing guidelines or ethics and AI policy which can then be explicitly referenced in your prompt.

  • Provide as much detail as possible in your prompt. For example instead of describing a car, you could describe a ‘sleek, midnight blue Tesla’. Refer to the Pencil templates for further examples of good use of detail

  • Be clear and specific. The more specific you are, the better the output will be.

  • If you need text organised in a specific way (e.g. bullet points) include the desired structure in your prompt.

  • Specify the tone and style (formal, conversational etc.)

  • Specify audience. If your text is aimed at a particular audience, include it in your prompt. Tailoring the content to the audience can enhance its effectiveness.

  • Use examples - if you have a particular style or format in mind, include examples in your prompt to help the model understand your expectations.

  • Iterate on prompts. If the first output isn’t quite what you hoped for, adjust your prompt and experiment with wording - this can yield different results.

  • Optionally, limit the length of your text by specifying a word count.

  • Review and edit - feel free to ask the model to refine or expand on specific parts, or adjust the structure.

Did this answer your question?