How to use AI to write a cover letter

Can AI write a good cover letter?

Yes — as a first draft — if you give it three things: the job description, your real achievements, and a sense of how you actually sound. No — if you paste "write me a cover letter" and hope for magic. The quality of an AI cover letter is decided almost entirely by the quality of what you feed it.

Think of the model as a fast, capable writer who knows nothing about you or the role until you tell it. Hand it a job description and three specific things you've done, and it will assemble a tight, tailored draft in under a minute. Give it nothing, and it fills the gap with the average of every cover letter on the internet — which is exactly the generic mush hiring managers have learned to skim past.

The step-by-step method

Here's the repeatable method. It takes about five minutes and works for any role.

Give it the job

Paste the full job description and the company name into your prompt. This is the single biggest lever. The model can only tailor the letter to the role if it can see the role — the responsibilities, the required skills, the language the company uses. Skip this step and every letter you generate will be interchangeable.

Give it your evidence

List two or three real achievements, with specifics and numbers. Not "I'm a strong marketer" but "grew a newsletter from 4,000 to 22,000 subscribers in eight months." Concrete outcomes are what the model turns into persuasive sentences. It cannot invent these — they have to come from you, and they're what separate your letter from everyone else's.

Set the tone and length

Tell it how long (250 words is plenty) and how to sound. "Warm and direct, first person, plain English, no buzzwords" beats leaving it to default into corporate stiffness. If you have a sample of your own writing, paste a few sentences so it can match your rhythm.

Ask for a draft, not a final

Frame the output as a starting point. Ask for a draft you'll edit, and consider requesting two different openings so you can pick the stronger one. You're using the model to get past the blank page, not to hand in its first attempt.

Edit ruthlessly

This is the step most people skip, and it's the one that matters. Read the draft out loud. Cut every phrase you'd never actually say. Delete clichés, tighten the opener, and make sure every claim is backed by a real specific. Five minutes of editing is the difference between a letter that sounds like you and one that sounds like a robot.

A strong starting prompt looks like this: "You're helping me write a cover letter for a Product Marketing Manager role at [Company]. Here's the job description: [paste it]. Here are three things I've actually done: (1) grew a SaaS newsletter from 4,000 to 22,000 subscribers in eight months; (2) led the launch of a new $40/month tier that hit 1,200 sign-ups in its first quarter; (3) rebuilt the onboarding email sequence and lifted activation 18%. Write a 250-word draft in a warm, direct, first-person voice — no buzzwords, no 'I am passionate about.' Open with a specific reason I'm a fit for this role, not a summary of my resume."

The generic-AI tells to cut

Hiring managers read hundreds of these. They've learned to spot AI autopilot in seconds. Here's what gives it away — cut all of it:

• Buzzwords and filler: "passionate," "leverage synergies," "dynamic self-starter," "think outside the box." These say nothing. Replace them with what you actually did.
• No specifics: a letter that never names a number, a project, or a real result is a letter about nobody. Specifics are the whole point.
• Over-formal openers: "I am writing to express my keen interest in the position of…" No human talks like this. Open with a reason you're a fit, or a genuine hook about the company.
• Claims with no backing: "I'm an excellent communicator" means nothing on its own. Show it — reference the thing you communicated and what happened.

The honest truth: AI makes it easier than ever to produce a mediocre cover letter, because the default output is confident, grammatical, and completely generic. The candidates who win are the ones who use AI for the drafting speed and then do the specific, human editing the model can't.

Do it faster next time

Once you've written one good prompt, save it. The scaffolding — paste the job, list achievements, set tone, ask for a draft — doesn't change between applications. Swap in the new job description and your most relevant wins, and you've got a tailored draft for the next role in two minutes instead of an evening.

That's the real payoff: the skill is reusable. Getting good at this kind of prompting — giving the model context, evidence, and constraints — pays off far beyond cover letters. It's the same technique that makes AI useful for writing a resume, drafting emails, and prepping for interviews, and it's worth learning to write a prompt properly. Iro AI turns prompting into short daily practice: you write real prompts, get instant feedback, and level up the exact skill that makes AI produce work worth sending. It's free to start on iOS or runs in your browser. Start with a free AI IQ test to see where you stand, or drill on real prompt exercises.