You’ve crafted a cover letter for a job application, but something feels off. Perhaps your writing seems a bit dull—after all, it’s tough to talk about yourself convincingly—or the letter appears unfocused. Instead of re-editing it multiple times, you can leverage ChatGPT with just a right-click to assist.
However, I don’t recommend letting ChatGPT generate the entire text for you. Even if you try to adapt AI models like Claude or Gemini to match your voice, their peculiar quirks—such as their tendency to emphasize differences (“it’s not this, it’s that!”)—often sneak in. That’s why an AI-generated cover letter might do more harm than good, landing your application in the reject pile.
A smarter approach is to use ChatGPT as a writing coach rather than a ghostwriter. This way, you benefit from its analytical skills without sacrificing your authenticity.
Here’s an effective prompt to try:
Here’s my [email / post / message / draft]. Don’t rewrite it. Examine it thoughtfully as an editor and tell me: what areas are weak, unclear, or likely to bore or confuse a reader? Be honest and direct — I want your genuine feedback, not a polished version.
This prompt serves two main purposes. First, it prevents the AI from simply polishing your draft to conform to generic standards, which can often result in bland, cookie-cutter writing filled with overly smooth Edits like generic punctuation or “it’s not this, it’s that” phrasing.
Second, it discourages the AI from flattering you and instead prompts it to identify specific issues—whether it’s vague language, meandering points, flawed structure, or dull passages.
For example, I recently used this prompt on a draft and received a brutally honest critique from ChatGPT’s GPT-5.5 “Thinking” model:
While your main idea is timely and valuable, the article drags on before reaching it. The initial setup overwhelms the practical message. Six paragraphs focus on “what happened / what might have happened” before delivering the lesson, which disrupts the flow.
Ouch, but accurate. ChatGPT even pointed out that one paragraph was “dense and bureaucratic,” and highlighted a logical flaw: “If [Claude] Fable is unnecessary for most users, why should they care that it’s gone?” This type of blunt but helpful criticism is often more useful than a polished rewrite.
This “critique, don’t rewrite” strategy is popular precisely because it’s so effective. The feedback from this prompt can be quite sharp, which is why you might follow it up with a question like:
Tell me which single issue, if fixed, would most improve the clarity or impact for the reader.
This additional prompt helps you focus on the most critical problem, making the feedback more manageable and actionable rather than overwhelming with a laundry list of flaws.