You're Using AI Wrong. Here Is the Quick Fix.

Susan Frew • May 21, 2026

Share this article

If AI keeps giving you mediocre output, the problem is almost certainly your prompt, not the tool.

Quick take

  • This issue breaks down five specific upgrades you can apply to any prompt right now: assign it a role, describe your audience, show it an example, tell it what to avoid, and ask for multiple options.
  • There is also one dead-simple four-step framework called RACE that ties it all together in 30 seconds flat.
  • Your action step: find your last disappointing AI output, apply one upgrade, and run it again.
  • Grab the free Prompt Upgrade Toolkit and get 20 ready-to-use before-and-after examples for the tasks you do every day.

Your AI was not broken. Your prompt was. Now you know how to fix it.

I am going to say something that might sting just a little. If you have been using AI and thinking "this thing is kind of mediocre," the problem is probably not the AI. It is what you are telling the AI to do. And the fix takes about five minutes to learn.

Most people type a quick request, get a generic response, shrug, and move on. That is like handing a brilliant new hire a sticky note that says "write a thing" and being disappointed with the result. The tool is only as good as the direction you give it. And once you understand how to give better direction, everything changes.

The Real Reason AI Gives You Generic Garbage

AI models are trained to be helpful to everyone. That means when you give a vague prompt, they default to the most average, middle-of-the-road response possible. They have no idea who you are, who you are talking to, what you have already tried, or what "good" looks like for your specific situation. So they guess. And their guess is fine for nobody in particular.

The fix is not complicated. You just have to stop writing prompts like search queries and start writing them like briefs. Give the AI context. Give it a role. Tell it who the audience is. Tell it what you want the output to look like. That shift alone will transform what comes back.

5 Prompt Upgrades You Can Use Right Now

These are real before-and-after examples. The difference in output quality between the two versions is not small. It is the difference between something you delete and something you actually send.

Upgrade 1: Give it a role. Before: "Write a follow-up email to a client." After: "You are a senior account manager at a consulting firm. Write a warm, professional follow-up email to a client who attended our strategy session last Tuesday. The goal is to recap the key decisions, confirm next steps, and leave them feeling confident about our partnership."

When you tell the AI who it is, it stops writing like a robot and starts writing like the person you described. Try it once and you will never skip this step again.

Upgrade 2: Describe your audience. Before: "Write a LinkedIn post about AI for business." After: "Write a LinkedIn post aimed at small business owners who are curious about AI but feel overwhelmed and slightly skeptical. The tone should be conversational and encouraging. Lead with a relatable pain point and end with one specific action they can take today."

The more specific you are about who you are talking to, the more the AI can tailor the language, the examples, and the emotional tone. Generic audience in, generic content out. Specific audience in, content that actually resonates.

Upgrade 3: Show it an example. Before: "Write a proposal introduction." After: "Here is an example of a proposal introduction I have written before that I really like: [paste your example]. Write a new introduction for a proposal to a manufacturing company using the same tone, length, and structure."

This is called few-shot prompting, and it is one of the most powerful tricks in the book. Instead of describing what you want, you show it. The AI reverse-engineers your style and replicates it. This is how you get AI output that actually sounds like you.

Upgrade 4: Tell it what to avoid. Before: "Write a bio for my speaker page." After: "Write a speaker bio for my website. Do not use corporate buzzwords. Do not start with my job title. Avoid phrases like 'thought leader,' 'passionate about,' or 'leveraging synergies.' Make it conversational, confident, and specific to my real experience."

Negative instructions are underused and wildly effective. If there are things you hate seeing in AI output, just say so. The model wants to please you. Tell it what not to do and it will stop doing it.

Upgrade 5: Ask for options, not just one answer. Before: "Write a subject line for my newsletter." After: "Write 5 subject line options for this newsletter. Give me one that leads with a question, one that leads with a surprising stat, one that is slightly provocative, one that is warm and encouraging, and one that uses humor. Keep each under 10 words."

When you ask for one thing, you get one answer and you are stuck with it. When you ask for five with specific parameters, you get a menu. You mix, match, steal the best parts, and end up with something better than any single option would have been.

The One Framework That Ties It All Together: RACE

If you want a single structure to remember, use this. It is called RACE and it takes 30 seconds to apply to any prompt.

  • R: Role. Who is the AI being right now?
  • A: Action. What exactly do you want it to do?
  • C: Context. What does it need to know to do it well?
  • E: Execute. What format, length, or style should the output take?

Before you hit send on any prompt, run it through RACE for 30 seconds. You will be amazed how much sharper your results get, every single time.

Your Action Step

Go find the last AI output you were disappointed with. Maybe it was a draft email, a social post, a proposal section. Open it back up, apply one of the five upgrades above, and run it again. Just one upgrade. See what comes back. I will bet you close this tab feeling very differently about AI than you did when you opened it.

This Is What Recoded Looks Like

Transforming your business DNA with AI does not start with the fanciest tool or the biggest budget. It starts with learning how to communicate with the technology you already have access to. Prompting well is a skill. And like every business skill worth having, it gets easier and faster the more you practice it.

When you get good at this, AI stops being a novelty you poke at occasionally and starts being the unfair advantage you reach for every single day. That is the shift. That is what competing on awesome actually looks like in practice.

Better prompts, better results, better business. You were one sentence away from it the whole time.

Want to skip the trial and error? Grab your free Prompt Upgrade Toolkit: 20 Before and After Prompts for Business Owners and start getting better AI results on your very next request.

Get the Prompt Upgrade Toolkit

Find more at susanfrewspeaks.com.

— Susan


Recent Posts

Susan Frew video on five tech outages in eight days
By Susan Frew • November 19, 2026
Microsoft 365, the FAA, Chase and Apple all went down in eight days, and none were hacks. Here's the small business lesson and a free AI prompt for your Plan B.
By Susan Frew • November 12, 2026
Your next customer is probably already in your files. Export your list, let AI find who went quiet, and send a personal note in 15 minutes. Free prompts inside.
Title card: AI Is a Great Stylist. It's a Lousy Friend.
By Susan Frew • October 29, 2026
My AI gave me a 10 out of 10 on an orange jacket. Here's why that worries me, what the JED Foundation research shows, and a free 20-minute AI stylist handout.
Title card: Four AI Assistants Are Already Living Inside Your Browser
By Susan Frew • October 22, 2026
ChatGPT, Claude, Gemini, and Copilot already live in your browser. Here is how to use them on your open tabs, plus official links and a free cheat sheet.
Title card: This Week's AI Drops, Sorted by What You Pay For
By Susan Frew • October 8, 2026
OpenAI, Google, and Anthropic all shipped in eight days. Here is who gets Dots, Gemini 4 Argon, and Claude Sonnet 5.5 on each plan, plus a free cheat sheet.
Title card: I Hired 11 Employees. None of Them Eat Lunch.
By Susan Frew • September 30, 2026
I built an AI org chart with 11 named agents. Two are already live, and a human still checks everything. Here's the six-line job description and a free worksheet.
Title card: The Facts Behind the September 2026 AI Panic
By Susan Frew • September 17, 2026
A researcher resigned from Anthropic, a DeepMind researcher followed, and Dario Amodei published a new essay. Here are the facts, sources, and what I would do.
Title card: Set It Once. Your AI Remembers Everywhere.
By Susan Frew • September 11, 2026
Google expanded Gemini custom instructions across Workspace. Tell it who you are once, and it uses your voice in Gmail, Docs, Drive, Chat, Sheets, and Slides.
Title card: I Let Claude Click Around Without Me
By Susan Frew • September 3, 2026
Claude in Chrome can now act on its own, with a safety check on every click. Here's how to try it carefully, flip the toggle, and keep control of the keys.
Title card: One Click, One Video, Zero Extra Software
By Susan Frew • August 27, 2026
Google built a Record button into Slides through Google Vids. Talk through a deck, fix flubs in the transcript, and send a shareable video with no extra software.
Show More