7 Prompting Secrets That Will Transform How You Use AI

 

Stop Getting Terrible Results and Start Getting Genius-Level Output

Here's what happens at nearly every conference where I speak about AI:

Someone raises their hand: "I tried ChatGPT and it gave me absolute rubbish. AI just doesn't work for my industry."

Then I ask to see their prompt.

Nine times out of ten, it's something like: "Write me a blog post about leadership."

The problem isn't AI. The problem is they're speaking to a genius like it's a mind reader.

I've spent three years testing AI tools across dozens of business scenarios—from sales proposals in Japan to keynote development in Dubai. The difference between mediocre AI output and genuinely brilliant results comes down to how you ask the question.

1) Give AI a Role (And Watch the Magic Happen)

Most people start with "Write me..." That's like walking into a party and immediately asking someone for a favour.

Instead, tell AI who it should be.

Weak: "Write an email to a difficult client."

Strong: "You are an experienced customer success manager with 15 years in B2B software, known for turning around difficult relationships through empathy and clear communication. Write an email to a client threatening to cancel because of a recent service issue."

See the difference? The specificity makes all the difference.

2) Show, Don't Just Tell

Want AI to write in a specific style? Don't describe it—show it an example.

I learned this after getting frustrated with AI content that sounded nothing like my voice. Then I started including examples of my own writing in prompts. Everything changed.

Every time I need content in my style, I provide a link to www.frankfurness.com/articles so it can check my writing and match my rather crazy sense of humor.

I use Claude to do my articles, it is brilliant, better than any other AI app.

Frank's Quick Tip: Create a "style file" with your three best pieces of content. Use snippets whenever you need AI to match your voice.

3) Use the "Think Step-by-Step" Trick

This sounds too simple to work, but it's genuinely transformative.

When you need AI to solve something complex, add: "Think through this step-by-step" or "Let's work through this methodically."

Why? AI models literally perform better when prompted to show their reasoning.

I was working with a sales team in Kuala Lumpur using AI to analyse deal risks. Initial prompts got surface-level responses.

When they added "Analyze this step-by-step, considering each risk factor individually," the quality jumped dramatically.

Use this for strategy development, problem-solving, complex analysis—any situation where you need thorough thinking rather than quick answers.

4) The Power of Constraints

Counterintuitively, limiting AI makes it better.

Instead of "ideas for improving customer retention," try: "Give me exactly three customer retention strategies that require no additional budget, can be implemented in under two weeks, and specifically target our mid-tier B2B clients. The name of our business is Frank Furness and the website www.frankfurness.com"

Constraints force strategic thinking rather than generic churn.

I use them constantly:
• "In exactly 150 words..."
• "Using only strategies that worked in the UK market..."
• "Avoiding any corporate jargon..."
• "Suitable for a technical audience with no marketing background..."

The tighter your constraints, the more focused and useful the output.

5) Build Context Like You're Briefing a New Team Member

AI doesn't know your business, industry nuances, or specific challenges unless you tell it. More context upfront means better output.

If you hired someone brilliant but new to your company, you wouldn't just say "improve our sales process" and expect magic. You'd explain your current process, where it breaks down, what you've tried, and what success looks like.

Do the same with AI.

Recent prompt I used: "I'm a B2B sales speaker working with offshore financial services companies in the Middle East. They're traditional, relationship-focused, and sceptical of new technology.

Sales cycles are 6-18 weeks with multiple decision-makers. I need to explain why AI-enhanced sales processes won't replace relationship building but will make relationships stronger. Draft an outline for a 45-minute presentation."

That gets infinitely better output than "Write a presentation about AI in sales."

6) Ask AI to Analyse Your Own Prompting

Here's a secret most people never discover: You can ask AI to teach you how to use AI better.

Try this: "I've been using AI for [your use case]. Here are examples of prompts I typically use: [paste 3-5 actual prompts]. Based on these, what am I doing well, and what could I improve? Give me specific suggestions."

Last week I asked ChatGPT to analyse all my prompts from when I first signed up. I was astounded.

AI will spot patterns you didn't realise existed, suggest improvements you hadn't considered, and teach you techniques specific to how you work.

I discovered I was being too vague in my context-setting. Changed my approach, and quality improved by at least 30%.

7) Use the Iterative Refinement Method

The biggest mistake? Expecting perfection on the first try.

The pros know: AI is a conversation, not a vending machine.

Your first prompt gets you 60-70% there. Then you refine: "That's good, but make it more conversational," or "Keep the structure but add healthcare examples," or "Too formal—write it like you're explaining to a colleague over coffee."

I rarely use AI's first output unchanged. I treat it like working with a talented junior colleague: they give me a solid draft, I provide direction, they refine, we iterate until it's right.

Here in Great Yarmouth by the coast or at my condo in Orlando, I write dozens of articles and newsletters every month.

None of them are AI's first attempts. They're all the result of this back-and-forth refinement—and honestly, that's where real quality comes from.

Right now, I have taken 2 days to write and submit 9 articles (my exact method in my webinar coming up)

The Bottom Line

AI is probably the most powerful business tool you've ever had access to. But like any powerful tool, the results depend entirely on how you use it.

These seven secrets aren't really secrets—they're techniques anyone can learn and apply immediately.

The difference between people who think "AI doesn't work" and people getting genuinely transformative results usually comes down to mastering these fundamentals.

Start with one. Maybe it's giving AI a role, or maybe it's asking it to analyse your prompting style. Master that, then add another. Within a month, you'll be getting output that makes colleagues ask, "How did you do that so quickly?"

And here's the thing: These prompting skills are only going to become more valuable. As AI tools evolve, the people who know how to communicate effectively with them will have an increasingly unfair advantage.

So, the question is: Will you be one of them?

Ready to transform your team's AI capabilities?

I work with organisations in 70+ countries who are committed to staying ahead. Visit frankfurness.com or contact me at frank@frankfurness.com | +44 7711 672888 | +1 407 588 9714

Why Training Your People on AI Isn't Optional Anymore

 

Let me ask you something: When was the last time your company rolled out new software and just assumed everyone would figure it out on their own?

Exactly. You wouldn't dream of it.

Yet that's precisely what most businesses are doing with AI right now. They're giving employees access to ChatGPT or whatever flavour-of-the-month they've subscribed to, then wondering why adoption is patchy and results are underwhelming.

I've trained teams in over 70 countries, and I can tell you this with absolute certainty: Access without training is just expensive confusion.

The businesses winning right now aren't the ones with the best AI tools. They're the ones whose people actually know how to use them properly.

 
1) The £50,000 Question Nobody's Asking

Three months ago in London, I had this conversation with a CEO of a 200-person firm:

Him: "We've given everyone AI tools. Cost us £50,000 a year. Usage is terrible."

Me: "How much did you spend training them?"

Him: (long pause) "We sent an email and did a bit of internal training."

There's your problem.

You wouldn't hire a salesperson without training them on your sales process. You wouldn't buy expensive machinery without training operators. But somehow, with AI, businesses think access equals competence.

It doesn't.

The companies that invest in proper AI training see 3-5x higher adoption rates. The ones that don't are burning money on subscriptions nobody uses effectively.

2) The Productivity Mirage

Most employees use AI like a slightly smarter Google. They ask basic questions, get basic answers, then declare "AI doesn't work for my job."

But when you train people on effective prompting, context setting, and AI-assisted workflows? Magic happens.

I worked with a marketing team in Atlanta using AI to "write social posts." Generic rubbish. Totally unusable.

Two hours of training later, they were generating 15 message variations, analyzing competitor content, brainstorming campaigns, and creating first-draft copy that actually sounded like their brand.

Same tool. Different results. The only variable? Training.

The art of prompting and how to use it is one of the most essential skills required right now. I think they will be teaching it in Universities in the future, if not already.

Frank's Quick Tip: Don't train people on "AI." Train them on "AI for your specific role." A finance person needs different skills than a marketing person. Generic training produces generic results.

3) The "Shadow AI" Problem (That Should Keep You Awake)

Your employees are already using AI. They're just doing it badly, without oversight, and possibly creating massive compliance or security risks.

When I audit companies, I consistently find:

• People pasting confidential information into public AI tools

• Employees using AI-generated content without checking accuracy
• Teams making decisions based on AI outputs they don't understand
• Complete inconsistency across departments

 Untrained AI use isn't just ineffective, it's dangerous.

Proper training teaches people how to use AI safely, what data they can and cannot share, how to verify outputs, and when AI is appropriate (or inappropriate) for different tasks.

4) The Competitive Time Bomb

While you're debating whether AI training is worth the investment, your competitors are already doing it.

Last month in Dubai, a retail company shared their results after comprehensive AI training:

• Customer service response times down 60%, satisfaction up
• Product team research time cut by 40%
• Marketing team doubled content output without hiring anyone

Meanwhile, their nearest competitor is still "evaluating AI strategy."

Guess who's winning market share?

This is happening right now. Businesses that train their people on AI are moving faster, serving customers better, and innovating quicker. The ones that don't are falling behind in real-time.

 
5) The ROI Is Ridiculous

Let's do some simple maths (yes, "maths" with an 's'—I'm British).

Invest £500 per employee in comprehensive AI training. That's typically one or two full days.

If that training helps each person save just two hours per week through better AI usage, you've got a 10-hour monthly return. At an average loaded labour cost of £50/hour, that's £500 saved per month per person.

Your investment pays for itself in one month. Everything after is pure profit.

I've seen companies achieve far better than two hours weekly savings. Some teams are saving 10-15 hours per person per week. The ROI isn't just good—it's spectacular.

6) What Training Actually Does (That Access Never Will)

Training creates a culture of intelligent AI adoption.

When everyone understands AI's capabilities and limitations, you get:

• Teams collaborating on AI-enhanced workflows
• People sharing effective prompts and techniques
• Innovation bubbling up from unexpected places
• Natural champions who help others improve

I worked with a manufacturing company in Spain where AI training sparked an internal innovation competition. Employees from warehouse to executive suite started sharing clever AI applications they'd developed.

Production efficiency up 18%. Employee engagement up. Turnover down.

None of that happens with just access.

 
The Bottom Line

Giving your employees access to AI without proper training is like giving them a Ferrari and expecting them to win races without driving lessons.

They might eventually figure out how to move forward, but they'll never unlock the real performance—and there's a decent chance they'll crash along the way.

The businesses that will dominate the next decade aren't the ones with the most sophisticated AI tools. They're the ones whose people know how to use AI effectively, safely, and strategically.

Training isn't an expense. It's a competitive advantage you either choose to build or choose to surrender to someone else.

From where I'm sitting—having trained thousands of professionals across every continent—the companies investing in AI training right now are the ones I'd bet on. The ones still "thinking about it" are the ones I'd be worried about if I were their shareholders.

The technology is here. The tools are available. The only question left is: Are your people actually ready to use them?

Want to transform your team from AI-curious to AI-capable?

I've developed training programs that take people from zero to genuinely effective in record time—no technical background required. Visit frankfurness.com to see what's possible for your business.