AU & NZ
SG
MY
US
IE
What is AI prompting?
AI prompting is the way people give clear instructions to AI tools so they produce useful, accurate outputs for specific business tasks.
In simple terms: Your input determines the output.
You do not need specialist knowledge. You only need to explain:
-
What you want
-
Why you want it
-
Who the output is for
-
How the output should look or sound
Mastering this skill helps teams improve clarity, accuracy and speed across a wide range of tasks.
Why AI prompting matters for organisations
Prompting is one of the most accessible ways to begin working with AI because:
-
Anyone can learn it
-
It does not require coding or technical skills
-
It is low cost to use and scale
-
It improves accuracy and quality of outputs
-
It reduces time spent on manual work
-
It supports consistent and responsible AI use
Prompting is often the most practical place to start for organisations beginning their AI journey.
How teams use AI prompting at work
Finance Teams
-
Summarising financial reports
-
Creating first drafts of commentary for board packs
-
Explaining variances
-
Preparing supplier communications
-
Trend analysis
HR Teams
-
Drafting job descriptions
-
Drafting CV summaries against a job description for a recruiter to review
-
Writing internal communications
-
Developing interview questions
-
Preparing policy summaries
Marketing and Communications Teams
-
Writing content for campaigns
-
Drafting social media posts
-
Simplifying technical product information
-
Summarising customer reviews
-
Preparing event briefing notes
These are just some of the everyday tasks where prompting can save time and improve consistency.
How to write effective prompts
To keep prompting simple and consistent, consider the ACE prompting framework. This framework helps teams write clearer AI prompts that produce more relevant responses.
A — Audience
Who is the output for? What do they need to understand?
C — Context
What background or purpose does the AI need to know?
E — Expectations
What should the output look like? Include tone, structure and length.
See the difference: a vague prompt vs a structured one
Here is the same task, written two ways. The contrast shows what “clear, structured prompts” actually means in practice.
1. Vague prompt:
“Write something for our team about the new £50 limit expenses policy.”
Result: a generic, off-tone paragraph that probably needs rewriting.
2. Structured prompt (using the ACE framework introduced above):
“Audience: All employees in our 120-person UK business.
Context: We are introducing a new expenses policy on 1 June. The main change is a £50 limit on client lunches and a requirement to submit receipts within 14 days. Attached is the revised expense policy.
Expectations: Draft a 150-word internal email from the Finance Director. Friendly, professional tone. Lead with what is changing, then why, then what employees need to do. End with a single sentence on where to ask questions.”
Practical prompting techniques anyone can use
These simple techniques include practical AI prompting examples that can help improve output quality immediately.
1. Provide an example
Show the AI a structure or style you want it to follow.
2. Refine incrementally
Review the result and adjust the prompt rather than starting again.
3. Request multiple options
Ask for two or three variations to compare approaches.
4. Ask the AI to clarify unclear expectations
For more complex pieces of work, end your prompt with:
“If anything is unclear, ask any questions you need before completing the task.”
Three worked examples you can adapt
Below are three prompts that show the ACE framework on everyday tasks.
Example 1: Drafting a job description
The task: you need a job description for a new hire and you want it to match how your existing job description’s read - not a generic AI-flavoured version.
“I have attached two of our existing job descriptions. These show our house style: section order, tone of voice, and the way we describe the company.
Audience: External candidates browsing job boards, often on a mobile.
Context: We need a job description for a Finance Manager. The role manages a team of three, owns the monthly management accounts, runs the annual budget process, and is the main finance contact for the senior leadership team. We are looking for a qualified accountant (ACA, ACCA or CIMA) with two or three years of post-qualification experience.
Expectations: Draft the new job description in the same structure and tone as the attached examples. Around 400 words. Easy to scan on mobile. Do not invent benefits or perks - leave that section as a placeholder for me to fill in.”
Why it works: showing the AI an example of your house style is one of the highest-leverage moves in prompting. The output sounds like you, not like generic AI copy.
Example 2: A seasonal promotion campaign
The task: you are running a summer promotion and want first drafts of every customer-facing asset in one go.
“We are running a summer promotion and I need first drafts of the campaign assets.
Audience: Existing customers and warm prospects on our mailing list. UK consumers, mixed ages.
Context: The offer is 20% off all bookings made in June for stays or visits during July and August. The promotion runs from 1 June to 30 June. The discount code is SUMMER20. Standard terms apply - not valid with other offers, subject to availability.
Expectations: Review our website <insert website> to understand our tone of voice, and craft the following, in this order:
1. A short email to our customer list (around 150 words) announcing the offer.
2. Three social media posts - one announcing the offer, one highlighting a specific reason to book, one final reminder in the last week.
3. A 60-word website banner with a clear call to action.
4. A short SMS message (under 160 characters) for customers who have opted in to texts.
Tone: warm, friendly, not pushy. British English. Every asset must include the discount code SUMMER20 and the deadline of 30 June.”
Why it works: The AI assistant can use web search to understand your tone of voice. One prompt produces a coherent set of assets all pointing at the same offer, with the tone, dates and offer code consistent across every channel.
Example 3: A monthly sales summary
The task: you have a sales report exported from your finance or sales system and need a short written summary for a management meeting.
“Attached is our raw sales data for last month, exported from our finance system.
Audience: The management team at our monthly review. A mix of finance, operations and commercial - not all of them look at sales numbers regularly.
Context: We have a monthly target of £XXXk in sales. The same month last year came in at £YYYk. We launched a new product line in March and want to see how it is performing.
Expectations: A short written summary (under 250 words) covering:
(1) total sales for the month vs target and vs the same month last year
(2) the top three products by value
(3) how the new product line is performing
(4) one trend worth flagging
(5) two or three recommended actions.
Plain UK English, no jargon. If any column is unclear, ask before drafting.”
Why it works: the AI knows what the file contains, who the output is for, what “good” looks like (length, structure, tone), and what to do if something is ambiguous.
How prompting supports responsible AI use
Clear prompting helps organisations adopt AI safely because it:
-
Encourages clarity and transparency
-
Reduces inaccurate or misleading outputs
-
Supports consistent behaviour across teams
-
Reinforces internal AI usage guidelines
-
Helps employees understand how to use AI responsibly
Prompting is both a productivity tool and a governance tool.
Where prompting fits in your AI journey
Our ‘Your AI Journey’ hub includes four practical stages that help organisations like yours build capability at the right pace:
1. Getting started
2. Ready to implement
3. Scaling AI
4. Leading the way
Prompting plays an essential role in the early stages, particularly Getting started and Ready to implement. It helps teams understand how to work effectively with AI tools, build confidence and reduce risk before moving into more advanced or integrated use cases.
To find out where your organisation currently sits, you can take the AI Readiness Quiz, a quick way to assess your maturity level and identify your next steps.
For tailored support, practical tools and guidance matched to your stage, visit our ‘Your AI Journey’ hub, where you can explore resources designed to help you build safe, confident and productive AI adoption across your organisation.
Frequently Asked Questions
What is AI prompting in simple terms?
It means telling an AI tool what you want using natural language. Clear instructions lead to better results.
How does AI prompting improve productivity in organisations?
It speeds up tasks such as writing, summarising and analysis, reducing manual work across teams.
Do employees need technical skills?
No. Anyone can learn prompting. It is based on communication, not coding.
How do we ensure prompting is used safely?
Set clear guidelines and treat AI tools the same as any other system that handles your data. Under UK GDPR, do not paste personal data, customer information or commercially sensitive content into consumer or free-tier AI tools, where inputs may be used to train the underlying model. For this you should use a business or enterprise tier (for example Claude for Enterprise, Microsoft 365 Copilot, ChatGPT Business or equivalent) where the provider contractually commits not to train on your data. Always review outputs before using them, and use structured prompts so colleagues can see how a result was produced.
How detailed should my prompts be?
You do not need to write long prompts, but detail matters. Include enough context for the AI to understand the task, the audience, and what good output looks like. Most prompts benefit from two to four clear sentences.
Can prompting help reduce workload during busy periods?
Yes. Prompting can help create first drafts, summarise information and speed up repetitive communication, freeing up time for teams to focus on higher‑value tasks.
How can my organisation build prompting skills across teams?
Start by sharing simple frameworks like ACE (Audience, Context, Expectations), offer examples tailored to real‑world workflows, and encourage teams to practise using prompts for everyday tasks. Our ‘Your AI Journey’ hub also provides free practical guidance and tools matched to your AI maturity stage.
