Give yourself superpowers. 5 tricks for deep-thinking with AI.

Most conversations about AI in business still circle around speed. How many hours can we save on emails, reports or code? Useful questions, but not the most important ones.

The better question for SME leadership teams is this:

How can we use AI to think better, not just work faster?

When you treat AI as a shortcut, you get shortcut outcomes. When you treat it as a thinking partner, you get clarity, confidence and stronger decisions.

Recent research backs this up. Organisations that embed generative AI and automation into their core processes – not just bolt-on tools – are seeing around 2.5x higher revenue growth and 2.4x greater productivity than their peers. (Source: Accenture) The gains are not only in speed but also in better-quality judgment at scale.

This article shares a practical framework you can use immediately. Using a performance review as a case study, we will walk through five simple “tricks” that move you out of autopilot and into a more reflective, collaborative mode with AI.

Business takeaway: When leaders frame AI as augmentation rather than automation, it becomes a tool for sharper thinking, better decisions, and higher-value work.

Why SME Leaders Should Favour AI Augmentation Over Automation

Before the prompts, it helps to name the difference.

• AI automation is when a system acts on its own. It updates records, moves workflows forward or sends messages without a human making the final call.
• AI augmentation is when AI supports a person who remains in charge of the decision. The system summarises, compares, drafts or challenges, but you approve the outcome.

For most UK SMEs, augmentation is the better first move.

1. Risk and accountability stay clear
   When people remain in the loop, mistakes are easier to spot and correct. Automation requires deeper access to systems, stricter controls and heavier governance. That is achievable, but it raises the bar on readiness.
2. High-value work is rarely fully automatable
   Performance reviews, client conversations, hiring decisions and strategy work are all low-volume, high-consequence activities. They depend on context, trust and nuance. AI can inform and stretch your thinking, but it should not decide.
3. People trust augmentation more than replacement
   In one large survey, 75% of employees said they would be comfortable with AI drafting their performance review as long as their manager reviewed and adjusted it. The same research showed trust dropped sharply when AI acted alone. (Source: Betterworks AI Survey Report)
4. Augmentation compounds over time
   Each time your team uses AI to clarify a decision, challenge assumptions or unpack a problem, you gain institutional knowledge. That is very different from a bot firing off emails in the background.

Business takeaway: Focus AI investments on augmenting judgment in leadership, client and performance contexts first. You reduce risk, build trust and see clearer ROI than jumping straight to full automation.

How to Use AI for Strategic Thinking in SMEs

Most SMEs start with obvious “doing” use cases:

• Drafting emails and documents
• Summarising meetings and research
• Generating code snippets or marketing copy

Those are fine. They free up capacity. But the real leverage appears when you point AI at harder strategic questions:

• Where should we focus our limited investment over the next 12 months?
• What are three alternative pricing strategies for this product, and what are the trade-offs?
• How might our ideal client describe our strengths and weaknesses today?

Tools like ChatGPT, Copilot and Gemini work best when you treat them as conversational partners, not vending machines. MIT Sloan describes prompts as “conversation starters” that you build on iteratively rather than one-shot commands. (Source: MIT Sloan – Effective prompts for AI)

That mindset shift – from issuing instructions to holding a dialogue – is what unlocks deeper thinking.

Business takeaway: Use AI sessions as you would a working session with a colleague. Share context, ask it to probe and refine, and expect to go back and forth until your own thinking sharpens.

Five Practical Prompt Tricks for Deep Thinking With Generative AI

These five “tricks” are really question patterns. You can apply them to strategy, performance reviews, client issues or board papers.

Aim to use at least two of them in any AI-assisted task.

Trick 1: Change Perspective

Most of us default to our own point of view. AI makes borrowing from others cheap.

What to do
Ask the AI to respond as if it were a specific persona, with a specific stake in the decision. For example:

• “Act as a long-standing client who values reliability over inxovation.”
• “Respond as our finance director, looking at this proposal for the first time.”
• “Take the viewpoint of a new starter trying to understand how my role fits into the strategy.”

Example prompt

“You are a cautious but fair CFO in a UK SME. I am proposing a 20% increase in marketing spend focused on brand. Read this summary and tell me what would worry you and what evidence you would ask for before approving it.”

Business takeaway: Perspective-shifting surfaces blind spots quickly and helps leaders make decisions that hold up under scrutiny.

Trick 2: Unpack the Problem

Leaders are often handed symptoms, not root causes. AI can help you interrogate the why behind an issue.

What to do
Describe the situation, then explicitly ask “why” at least once. If the answer is still surface level, ask “what might be behind that” or “why else”. You are using AI as a structured thinking partner.

Example prompt

“Sales cycles have lengthened by 30% over the last six months. Here are the key data points. Help me generate at least five plausible root causes, grouped by internal, client and market factors. For each cause, suggest one low risk test we could run in the next 30 days.”

Business takeaway: Problem unpacking with AI stops you jumping straight to solutions and creates clearer, testable hypotheses.

Trick 3: Explore Widely

Our first three ideas are rarely our best ones. AI has no such limit.

What to do
Ask for a deliberately wide spread of options and make quantity part of the brief. Then you use your judgment to select and refine.

Example prompt

“We want to use AI for strategic thinking in our SME, not just content. List 15 specific situations where AI could help the leadership team make better decisions, grouped into finance, operations, marketing and people. Keep them practical and realistic for a 150 person company.”

Business takeaway: Wide exploration expands your option set, so you choose among strong alternatives rather than the first acceptable one.

Trick 4: Playact a Conversation

Performance reviews, client meetings and board updates are all conversations. AI is an ideal rehearsal space.

What to do
Ask the AI to roleplay a live dialogue with you. You speak as yourself; it speaks as the other party. Keep the exchanges short and realistic.

Example prompt

“Roleplay a conversation where I explain to a sceptical senior manager why we are prioritising AI augmentation over automation this year. Ask tough but fair questions about risk, cost and impact on their team.”

Used carefully, this is a safe way for managers to practise giving difficult feedback and for employees to test how they might respond.

Business takeaway: Playacting turns abstract worries into specific words, so real conversations feel calmer, clearer and more constructive.

Trick 5: Push Boundaries

AI can also help you see around corners.

What to do
Invite the AI to challenge your assumptions, stress-test your plan, or highlight ethical and client risks you may be underplaying.

Example prompt

“Here is our draft AI roadmap for the next 18 months. Identify the top five ways this could fail in practice for a UK SME, including cultural, regulatory and client risks. For each one, suggest a mitigation we can build in from the start.”

Business takeaway: Boundary-pushing turns AI into a constructive critic, making your plans more robust before they hit the real world.

Case Study: Using AI for Annual Performance Review Reflection

Let’s bring this to life with Emma, an imaginary client services manager in a UK financial services SME.

Emma has to write her annual self-assessment. She could use AI in automation mode – paste in some notes, ask “Write my performance review”, and paste the result into her HR system.

Instead, she chooses augmentation. She spends 30 minutes in a secure AI environment and uses the five tricks above.

Safety note: In real organisations, Emma should use an approved, enterprise grade tool such as Microsoft Copilot or an internal Azure OpenAI deployment, not a public chatbot, to protect client and employee data. Growcreate’s AI development services are designed around this type of secure Azure setup. (Source: Growcreate – AI development services)

Step 1: Change Perspective

Emma starts with this prompt:

“Based on these anonymised client comments and KPI summaries, how might my key clients describe my performance this year. Where would they say I am strongest and where would they wish I was bolder or more proactive.”

The AI reflects that clients probably see her as thoughtful, prepared, and protective of their interests, but may sometimes experience her as cautious when time-sensitive opportunities arise.

Step 2: Unpack the Problem

Emma then zooms in on the word “bolder”.

“My manager and two clients have used the word ‘bolder’ about my advice. Give me three different interpretations of that word in this context and what specific behaviours they might be noticing or missing.”

She learns that “bolder” could mean bringing her own recommendations forward earlier, challenging client assumptions more directly or quantifying the upside of action, not just the downside of risk.

Step 3: Explore Widely

Next, she asks:

“Generate ten practical ways I could show appropriate boldness with clients next year, without becoming reckless. Prioritise actions I can try in my next three meetings.”

The AI suggests actions such as sharing a clear point of view at the start of meetings, testing one “stretch” recommendation per quarter and agreeing in advance how much risk appetite each client really has.

Step 4: Playact the Conversation

Emma feels nervous about changing her style, so she rehearses.

“Roleplay a conversation where a long standing client pushes back on a bolder recommendation I make. Play the client realistically, including concerns about risk and internal politics. Help me practise how to respond calmly, using facts, stories and questions.”

After a short back and forth, she feels more prepared for the real thing.

Step 5: Push Boundaries

Finally, she zooms out.

“Looking at this whole reflection, what am I underestimating about my impact, and what risks am I downplaying. Suggest a development goal that would genuinely stretch me over the next year.”

The AI points out that her empathy and reliability are powerful assets and that becoming bolder is not about being louder, but about combining that empathy with clearer recommendations.

At the end of 30 minutes, Emma has:

• A sharper understanding of how clients and colleagues likely see her
• Language she can use in her self-assessment that feels authentic, not generic
• Concrete development goals that are meaningful to her and valuable to the business

When she later sits with her manager, they spend less time wordsmithing the form and more time discussing her growth.

Business takeaway: Using AI as a thinking partner before writing a performance review produces better insight, better conversations and, ultimately, better performance.

Turning Deep Thinking Prompts Into an AI Augmentation Habit for Your Team

One example is helpful. Real value appears when this becomes a shared habit across your leadership team.

Here is a simple way to start inside a UK SME.

1. Identify your high-value thinking moments
   List 5–10 recurring activities where judgment matters: annual planning, board reports, major client proposals, performance reviews, hiring, and pricing changes.
2. Create prompt playbooks, not policies alone
   Alongside your AI usage and data policies, create one-page prompt guides with question patterns like the five tricks above, tailored to your context.
3. Choose secure, integrated tools
   Use enterprise-grade platforms, such as Azure OpenAI integrated with Microsoft 365, so prompts and outputs stay inside your governance model. (Source: Growcreate – AI development services)
4. Measure more than hours saved
   Track changes in decision quality, confidence and fairness, not just time. In Betterworks’ research, almost three-quarters of employees believed generative AI could lead to more objective, unbiased performance evaluations when combined with human review. (Source: Betterworks – AI Survey Report) Those are leadership outcomes.
5. Normalise rehearsal and reflection
   Encourage managers to use AI to rehearse difficult conversations, explore scenarios and debrief major decisions. Over time, this builds a culture where reflection is part of the workflow, not an afterthought.

Business takeaway: A small library of shared prompts, wrapped in the right safeguards, can lift the quality of thinking across your whole leadership team.

Where Growcreate Fits If You Want AI Augmentation, Not Just Automation

If you are a CMO, CTO or SME leader, you do not just need another tool. You need a way to make AI a stable, trusted part of how your teams think and decide.

That is where Growcreate’s Azure first AI services come in.

Growcreate helps organisations design and build AI systems that sit alongside existing .NET and Azure platforms, with a strong emphasis on governance, security and responsible use. (Source: Growcreate – AI development services)

Typical ways we support leadership teams include:

• AI consulting for augmentation strategy – identifying where AI should support judgment, where it should automate and where it should stay out of the way. (Source: Growcreate – Services)
• AI development services – integrating Azure OpenAI, AI search and workflow automation into your existing systems so managers can use AI safely for research, reflection and decision support. (Source: Growcreate – AI development services)
• Practical AI at work guides – such as our 2026 guide to ChatGPT, Copilot, Gemini and agentic AI, which gives leaders a roadmap from experimentation to embedded capability. (Source: Growcreate – AI at work guide)

Because everything is engineered around Azure governance, ISO 27001 and GDPR, you can let teams experiment with prompts like the ones in this article without losing control of data or compliance. (Source: Growcreate – AI development services)

Business takeaway: Partnering with a specialist Azure and AI engineering team gives you a safer, faster route to making augmentation real inside your organisation.

Bringing It Back to You and Your Next Performance Review

The next time you face a blank performance form or a difficult strategic decision, notice your first instinct.

If it is “How can AI do this for me?”, pause.

Instead, try:

1. Change perspective
2. Unpack the problem
3. Explore widely
4. Playact a conversation
5. Push boundaries

In half an hour, you will know yourself, your clients and your options better than when you started.

Automation saves time. Augmentation creates value.

And value - not speed alone - is how AI gives you and your team genuine superpowers.