7 practical ways to improve your use of AI

You don't need to be a technical expert to improve your results, but you do need to be control conscious. Small changes in how you approach AI can make a significant difference.

1. Give AI the right context

One of the most common mistakes is asking a broad question and expecting a tailored answer.

For example, compare:

“Write me a business plan.”

With:

“I'm starting a small business (state what the business will do). My target customers are busy professionals (specific professions?). I have £x to start and can work y hours per week. Provide me with a 30/60/90-day business plan, prioritising activities that can generate revenue quickly. Highlight assumptions and ask me only questions that materially affect the plan.”

Because the second prompt gives AI a clearer understanding of your circumstances, objectives and limitations, the result is more likely to be relevant, realistic, and actionable.

The principle is simple: Context + objective + constraints = better output.

2. Tell AI what role to play

Depending on how the request is framed, AI will approach the question from different perspectives. For example, instructions such as:

• “Act as a critical business adviser. Identify weaknesses in my proposal.”

• “Act as an experienced teacher and explain this to someone with no prior knowledge.”

• “Act as a meticulous editor. Improve clarity and identify unsupported claims.”

• “Act as a customer service manager. Draft a professional response that protects the customer relationship without making promises we cannot keep.”

help AI to establish the perspective, expertise and tone you want the response to reflect.

NB - asking AI to act as an expert does not make its output expert verified. Important facts and recommendations need to be checked.

3. Ask for a process, not just an answer

For straightforward tasks, a direct question may be enough. For important business decisions, a structured approach is invariably more useful. Instead of asking:

“Which option should I choose?”

Use:

“Break this problem into its key components. Identify missing information, analyse the available options, explain the trade-offs and provide a concise recommendation. Clearly state the assumptions and risks behind the conclusion“

If you want to retain complete control over the decision, ask for a neutral comparison:

”Compare these options against cost, time, risk and potential benefits. Explain the trade-offs, but don't choose for me.”

This approach helps you understand the reasoning behind a recommendation rather than simply accepting an answer.

4. Make AI challenge your thinking

One of AI's most useful applications is helping you identify things you may have overlooked. It’s easy to use AI to reinforce an existing idea, but a better approach is to ask it to test that idea.

For example:

• What am I overlooking?

• Challenge my assumptions.

• Give me the strongest argument against my plan.

• What could cause this project to fail?

• What risks have I underestimated?

• What would an experienced professional notice that a beginner might miss?

This type of question will help to expose weaknesses, identify alternative approaches and improve the quality of your decisions. AI should not simply be a tool for confirming what you already believe. Used well, it’s a valuable way to examine your thinking from a different perspective.

5. Improve the result through iteration

Don't expect the first response to be the final answer, a more effective approach is to treat AI as part of a review and improvement process.

The Principle: Draft; Critique; Improve; Simplify; Verify - Then Finalise

For example, when improving business proposals, reports, customer communications, presentations and internal procedures

• Ask AI to review and improve the initial draft.

• Ask it to identify the weakest sections from a client and stakeholder perspective.

• Request a revised version that addresses those weaknesses by stressing the benefits of your proposal for the reader.

• Ask it to make the language more natural and concise.

• Check every detail.

• Rinse and repeat until you have an answer which meets your needs.

• Submit the proposal/blog post/presentation.

The same process can be applied for any question, for example when improving a CV:

• Ask AI to review and improve the initial draft.

• Ask it to identify the weakest sections from a recruiter's perspective.

• Request a revised version that addresses those weaknesses without inventing experience or qualifications.

• Ask it to make the language more natural and concise.

• Check every detail before submitting it.

It is vital that you do not eliminate human involvement. Use it to make each stage more productive.

6. Show AI what good looks like

Choose your words carefully.

“Professional”, “friendly” and “engaging” mean different things to different people. If you want AI to produce content that reflects your business's communication style, give it an example.

For instance:

“I want my emails to sound like this: [insert example]. Match the same level of warmth, clarity and directness, but do not copy the wording. Keep the message concise and appropriate for a business customer.”

This is particularly useful for businesses that want consistent communication across emails, proposals, reports and customer service responses. It also helps maintain a recognisable brand when several people use AI to create content.

7. Use AI for the repetitive 80% and keep human judgement where it matters

AI can be valuable for repetitive, time-consuming tasks that don't require your full attention.

Some examples:

• Summarising lengthy documents.

• Turning meeting notes into action plans.

• Drafting routine emails.

• Brainstorming ideas.

• Exploring unfamiliar subjects.

• Creating checklists and templates.

• Analysing spreadsheets and data.

• Preparing presentations.

• Comparing options.

• Proofreading documents.

• Producing drafts.

• Explaining technical concepts.

• Standardising repetitive processes.

• Preparing questions for meetings or interviews.

These applications can free up time for work that genuinely needs human expertise such as building relationships, solving complex problems, exercising judgement and making important decisions.

The key is to establish and implement effective controls, particularly when AI is handling confidential information, customer data or work with legal, financial, operational or reputational implications.