Prompting Craft
Get expert-level AI responses.
A good editor reads your draft differently from a copywriter does. Same with the AI: tell it which role to perform as, and the response narrows in a useful way. One sentence at the top of any prompt, about 2 minutes.
No credit card required. Cancel anytime.
Why not use the preset GPTs or expert personas built into AI tools?
ChatGPT has thousands of preset GPTs in its store: "Senior Brand Strategist," "Grant Editor," "Financial Journalist." Claude has Projects with custom instructions. Gemini has Gems. They all do versions of what this trail teaches, with the role pre-baked.
Pre-baked personas have one problem: they’re built for generic versions of the role, not for the way your discipline actually sounds. The "Senior Brand Strategist" GPT is averaged across what the internet thinks brand strategists do. Not the way someone in your industry, at your seniority, with your sensibilities, would actually read the draft.
Writing the role yourself, in one sentence ("Read this as a brand strategist who has worked with B2B SaaS for ten years"), gets you closer. You can shape the depth, the years of experience, the industry, the editorial sensibility. And change them in a single re-prompt. Works in any AI tool, takes 2 minutes to internalize.
Questions
- Is my data private?
- Yes. Your prompts and AI responses are deletable, exportable, and only visible to you. We never expose individual prompts to admins, teammates, or Joring employees.
- When does the role sentence actually change the response?
- Most when the topic has a distinct voice of expertise. Strategy, criticism, analysis, editing, code review. Less so for purely descriptive prompts ("summarize this article"). Use the role sentence when you want a perspective; skip it when you just want neutral information.
- How specific should the role be?
- Specific enough that two practitioners in the discipline would read the role the same way. "Brand strategist" is too vague. "Senior brand strategist who has worked with B2B SaaS for ten years" is right. "Mary at Pentagram on a Tuesday" is too specific. The model can’t access that.
- Can I stack roles?
- Yes. "Read this as a financial journalist used to writing for non-experts" is two roles in one sentence (journalist + audience), and it works well. Stacking three or more usually dilutes the signal.
- Does this work in ChatGPT, Claude, and Gemini?
- Yes, equally well. Role prompting is one of the most consistent techniques across modern LLMs.
- What happens after the 14-day trial?
- Continue on Pro to keep real-time coaching, guided trails, and your exportable AI record. Or stay on Free with limited features. The trail progress you’ve already made doesn’t disappear.
Get responses written from inside the right discipline.
No credit card required. Cancel anytime.