Stop Wasting ChatGPT: Prompting Techniques That Actually Work for Business

Stop Wasting ChatGPT: Prompting Techniques That Actually Work for Business

You've tried ChatGPT. You typed in a request, read the response, and thought: "That's... fine, I guess." Then you rewrote it yourself anyway.

Sound familiar? You're not alone. And the problem isn't the tool — it's how you're using it.

Most business owners approach AI the same way they'd use a search engine: ask a question, hope for a useful answer. But language models don't work like Google. They work more like a brilliant, inexperienced contractor. Give them vague instructions and you'll get vague results. Give them a clear brief and they'll surprise you.

That's what prompt engineering is — writing a clear brief. No code, no technical background required.

Why Generic Prompts Produce Generic Results

Large language models have processed an extraordinary amount of text. That's their strength — but it's also why they default to the average of everything they've seen. Ask them to "write a sales email" and you'll get something that sounds like every sales email you've ever ignored.

The fix isn't a smarter model. It's a smarter prompt.

Technique 1: Assign a Role

Before describing the task, tell the AI who it is for this specific job.

Generic: > "Write a proposal email."

Better: > "You're a senior account manager at a B2B software company. Write a proposal follow-up email to a mid-size retail chain that attended our product demo last week. They're evaluating three vendors and cost is their primary concern."

Now the model has a perspective, a relationship, and a context to write from. The result reads like it was written by someone who knows the client — not a template engine.

Technique 2: Specify Format and Constraints

The AI doesn't know your preferences unless you tell it. Shorter? Say so. Bullet points? Say so. No jargon? Say so.

Useful constraints to add: - "Max 150 words" - "Use bullet points, not prose" - "Plain English — no corporate buzzwords" - "Formal but warm — think professional services, not legal"

A single line of format guidance will do more for output quality than any prompt "hack" you'll find on social media.

Technique 3: Give Context About the Reader

AI writes for a generic audience unless you specify otherwise. Who is actually reading this?

• "This is for a CFO who is skeptical about ROI"
• "Our customers are mostly over 55 and not tech-savvy"
• "This goes to a legal team — they'll scrutinize every claim"

The more specific you are about the audience, the more the AI tailors language, argument structure, and tone accordingly.

Technique 4: Ask It to Think Step by Step

For analysis, decision memos, or structured documents, don't just ask for the output — describe the reasoning process you want the AI to follow.

Example: > "Compare these three CRM platforms. Do it in this order: 1) Summarize each product in two sentences. 2) Compare total cost of ownership over three years including onboarding. 3) List the biggest risk of each option. 4) Give a clear recommendation with reasoning."

This "chain-of-thought" approach dramatically reduces vague, wishy-washy analysis and produces something you can actually act on.

Technique 5: Iterate, Don't Restart

If the first result isn't quite right, don't start a new conversation. Just tell the AI what to change.

• "Make this shorter"
• "Change the tone — more direct, less apologetic"
• "The second paragraph doesn't land — rewrite it"
• "Add a specific example from the retail industry"

The model remembers your full conversation context. This iterative back-and-forth is where prompting becomes genuinely powerful — and why good results often come on the third or fourth exchange, not the first.

Three Ready-to-Use Templates

Email draft: > "You are [role] at a [industry] company. Write a [tone] email to [recipient]. Subject: [X]. Goal: get them to [Y]. Max [N] sentences."

Meeting summary: > "Here are notes from a team meeting. Create a structured summary with: 1) Key decisions made, 2) Next steps with owners, 3) Open items still to resolve. Format: bullet points."

Customer response: > "A customer is unhappy about [issue]. We [situation]. Write a professional, empathetic reply that acknowledges the problem, explains our solution, and rebuilds confidence."

The Competitive Edge Is Simpler Than You Think

Companies like Klarna, HubSpot, and Shopify haven't just deployed AI tools — they've invested in teaching their teams how to use them well. Not because their employees are poor writers, but because consistent, high-quality AI output requires consistent, well-structured input.

You don't need a formal training program to get started. You need five minutes, a willingness to be specific, and the understanding that a clear prompt produces a clear result.

At Sophera Consulting, we help businesses across the UK, US, and DACH region move beyond AI experimentation — embedding prompting practices and automation workflows that create measurable, lasting productivity gains.

---

Want to go beyond better prompts? Our free Automation Check identifies where AI automation creates real business impact — not just better email drafts. Book your 30-minute session and walk away with a concrete shortlist of opportunities.

→ Request your free Automation Check