Departments
AI use cases by department, with real figures
Every area of your company has repetitive tasks that AI already handles. Choose a department: you will see what can be automated, how much time is saved according to independent studies and how it is done step by step. I work with real cases from each department.
26 min
saved per person per day
15%
more issues resolved per hour
−40%
writing time
01 / 08
Management
Decide with information already summarised, compared and ready for the meeting.
In management, generative AI lets you decide with information already summarised, compared and ready for the meeting: 3.8× faster to catch up on a meeting, according to Microsoft Work Trend Index, 2023.
3.8×
faster to catch up on a meeting
29%
faster at searching, writing and summarising
26 min
saved per person per day (self-reported)
Summarising a meeting: 43 minutes without AI versus 11 minutes with AI.
Use cases
The problem
40-page reports nobody has time to read before the committee.
How AI solves it
- 1You upload the document.
- 2You ask for a summary of decisions, risks and key figures.
- 3You check the critical points in the original.
Example prompt
Summarise this report in one page for the management committee: decisions to make, risks, key figures and open questions. Give the page number for each data point.
What you gain
You walk into the meeting with the essentials and know where to check every figure.
Key figure for this department · Management: 3.8× faster to catch up on a meeting · Microsoft Work Trend Index, 2023
The problem
The agenda arrives late and the documents are scattered across several emails.
How AI solves it
- 1You gather the documents and the previous minutes.
- 2You ask for an agenda with objectives and timings.
- 3You adjust priorities and send it.
Example prompt
Using these documents and the previous minutes, prepare the agenda for the 60-minute committee: items, objective of each, time allotted, owner and pending decisions from the last meeting.
What you gain
Shorter meetings, with a clear objective for each item.
Key figure for this department · Management: 3.8× faster to catch up on a meeting · Microsoft Work Trend Index, 2023
The problem
You must choose between several options, each presented in a different format.
How AI solves it
- 1You describe the options and your criteria.
- 2You ask for a comparison table with pros and risks.
- 3You validate the assumptions before deciding.
Example prompt
Compare these three scenarios in a table by cost, timeline, risk and team impact. Flag the assumptions behind each and what information I am missing to decide.
What you gain
You decide on a common basis and see straight away what is missing.
Key figure for this department · Management: 3.8× faster to catch up on a meeting · Microsoft Work Trend Index, 2023
The problem
A major change to announce and no time to find the right tone.
How AI solves it
- 1You explain the change and who it affects.
- 2You ask for a clear message in the right tone.
- 3You personalise it and sign it.
Example prompt
Write an email to all staff announcing this change: what changes, why, from when and who to ask. Warm, direct tone, 150 words maximum.
What you gain
A clear message, sent on time, that heads off rumours.
Key figure for this department · Management: 3.8× faster to catch up on a meeting · Microsoft Work Trend Index, 2023
02 / 08
Marketing
Produce more content, of better quality, in less time.
In marketing, generative AI lets you produce more content, of better quality, in less time: −40% writing time, according to Science, 2023.
−40%
writing time
+18%
text quality
5–15%
of marketing spend in productivity value
Writing time: base 100 without AI versus 60 with AI.
Use cases
The problem
The creative meeting ends with lots of loose ideas and none prioritised.
How AI solves it
- 1You give the goal, audience and budget.
- 2You ask for ideas scored against your criteria.
- 3You pick the best two and develop them.
Example prompt
Suggest 10 campaign ideas to launch this product to SME managers in the UK. Score each from 1 to 5 on cost, originality and ease of measurement, and recommend the best two.
What you gain
You go from brainstorm to decision in the same session.
Key figure for this department · Marketing: −40% writing time · Science, 2023
The problem
The same message has to be rewritten for the website, LinkedIn and the newsletter.
How AI solves it
- 1You paste the base text.
- 2You ask for one version per channel, in its format.
- 3You review the tone and publish.
Example prompt
Adapt this text into three formats: a 120-word LinkedIn post with a hook, a newsletter paragraph, and a headline plus description for the website. Keep our warm, professional tone.
What you gain
A consistent message on every channel, without starting from scratch.
Key figure for this department · Marketing: −40% writing time · Science, 2023
The problem
Hundreds of reviews and comments that nobody reads in detail.
How AI solves it
- 1You export the comments.
- 2You ask for recurring themes with examples.
- 3You turn the themes into actions.
Example prompt
Analyse these customer reviews: group comments into themes, say how many fall in each, quote two verbatim examples per theme and suggest one concrete action for the top three.
What you gain
You know what your customers want, in their own words.
Key figure for this department · Marketing: −40% writing time · Science, 2023
The problem
The calendar is improvised every week and briefs arrive incomplete.
How AI solves it
- 1You set goals, channels and key dates.
- 2You ask for a monthly calendar with a brief per piece.
- 3You assign owners.
Example prompt
Create November's editorial calendar as a table: date, channel, topic, goal and call to action. For each piece add a three-line brief: key message, audience and format.
What you gain
A planned month and a team that knows what to produce.
Key figure for this department · Marketing: −40% writing time · Science, 2023
03 / 08
Sales
Less time on admin, more time with customers.
In sales, generative AI helps you spend less time on admin and more time with customers: 60% of a salesperson's time goes on tasks that are not selling, according to Salesforce, State of Sales.
60%
of a salesperson's time goes on tasks that are not selling
−40%
time to write proposals and emails
3–5%
of sales spend in productivity value
McKinsey, 2023 (estimate)
A salesperson's time: 60% on other tasks and 40% on selling.
Use cases
The problem
You arrive at the meeting without having reviewed the client's context.
How AI solves it
- 1You gather notes, the client's website and history.
- 2You ask for a prep sheet with questions.
- 3You review the sheet before going in.
Example prompt
Using this client information, prepare a one-page brief for my meeting: company context, likely priorities, five open questions to uncover needs and two risks to anticipate.
What you gain
You walk in prepared and ask better questions.
Key figure for this department · Sales: 60% of a salesperson's time goes on tasks that are not selling · Salesforce, State of Sales (survey)
The problem
Every proposal is written from scratch and follow-ups get forgotten.
How AI solves it
- 1You paste your meeting notes.
- 2You ask for the proposal in your usual structure.
- 3You check prices and terms before sending.
Example prompt
From my meeting notes, write a sales proposal: client need in their own words, proposed solution, three-stage plan and next steps. Add a follow-up email to send five days later.
What you gain
Proposals that speak to the client and follow-ups that go out on time.
Key figure for this department · Sales: 60% of a salesperson's time goes on tasks that are not selling · Salesforce, State of Sales (survey)
The problem
The CRM is full of data, but nobody draws conclusions from it.
How AI solves it
- 1You export your portfolio without personal data.
- 2You ask for segments and opportunities.
- 3You prioritise accounts to contact.
Example prompt
Analyse this anonymised customer export: group by sector and volume, identify accounts with no purchase in six months and suggest a contact order with the reason for each.
What you gain
You know who to call first and why.
Key figure for this department · Sales: 60% of a salesperson's time goes on tasks that are not selling · Salesforce, State of Sales (survey)
The problem
The client replies "it's expensive" and you don't know how to answer without losing the sale.
How AI solves it
- 1You paste the client's message.
- 2You ask for three replies with different approaches.
- 3You pick one and adapt it.
Example prompt
A client says our proposal is expensive. Write three short replies: one focused on value, one offering a reduced option and one asking a question to understand their budget. Respectful tone, no pressure.
What you gain
You reply quickly and with arguments, without improvising.
Key figure for this department · Sales: 60% of a salesperson's time goes on tasks that are not selling · Salesforce, State of Sales (survey)
04 / 08
Human Resources
Job ads, communications and hiring processes prepared in a fraction of the time.
In human resources, generative AI lets you prepare job ads, communications and hiring processes in a fraction of the time: ≈ 20% of the working week saved in recruiting teams using generative AI, according to LinkedIn, Future of Recruiting 2025.
≈ 20%
of the working week saved in recruiting teams using generative AI
−40%
time writing job ads and communications
Recruiting tasks: 5 days a week without AI versus 4 days with AI.
AI drafts, summarises and organises. The hiring decision is always human: the EU AI Act classifies recruitment systems as high-risk.
Use cases
The problem
Generic job ads that attract candidates who don't fit.
How AI solves it
- 1You describe the role and the team.
- 2You ask for a clear ad with real requirements.
- 3You check the language is inclusive.
Example prompt
Write a job ad for this role: mission in two sentences, five concrete responsibilities, must-have requirements separate from nice-to-haves, what we offer and how the process works. Inclusive language, no jargon.
What you gain
Applications that fit the role better from the start.
Key figure for this department · Human Resources: ≈ 20% of the working week saved in recruiting teams using generative AI · LinkedIn, Future of Recruiting 2025 (survey)
The problem
Internal policies that are long, outdated and hard to understand.
How AI solves it
- 1You upload the current policy.
- 2You ask for a clear version and an FAQ.
- 3You have the content validated by the right person.
Example prompt
Rewrite this remote-work policy in plain language: what is allowed, what isn't and how to request it. Add five FAQs with answers and flag any ambiguous point I should review.
What you gain
Fewer repeated questions and rules everyone understands.
Key figure for this department · Human Resources: ≈ 20% of the working week saved in recruiting teams using generative AI · LinkedIn, Future of Recruiting 2025 (survey)
The problem
HR announcements get little attention and raise questions.
How AI solves it
- 1You explain the news.
- 2You ask for a short message with the expected action.
- 3You post it on the right channel.
Example prompt
Write an intranet message about the new annual review campaign: what it is, deadlines, what each person must do and where to get help. 120 words maximum, friendly tone.
What you gain
Messages that get read and understood first time.
Key figure for this department · Human Resources: ≈ 20% of the working week saved in recruiting teams using generative AI · LinkedIn, Future of Recruiting 2025 (survey)
The problem
Improvised interviews, with different questions for each candidate.
How AI solves it
- 1You give the role profile.
- 2You ask for an interview guide with criteria.
- 3You use the same guide for every candidate.
Example prompt
Prepare a structured interview guide for this role: eight competency-based questions, what to look for in a good answer and a 1-to-4 rating scale. Include no personal questions.
What you gain
Comparable interviews and a better-informed human decision.
Key figure for this department · Human Resources: ≈ 20% of the working week saved in recruiting teams using generative AI · LinkedIn, Future of Recruiting 2025 (survey)
05 / 08
Administration
Documents, presentations and schedules handled in minutes.
In administration, generative AI lets you handle documents, presentations and schedules in minutes: 24 minutes saved writing a document, according to Government Digital Service (UK), 2025.
24 min
saved writing a document
19 min
saved creating a presentation
9 min
saved scheduling a meeting
Minutes saved: 24 on a document, 19 on a presentation and 9 scheduling a meeting.
Use cases
The problem
The monthly report is rebuilt by hand every month with the same sections.
How AI solves it
- 1You paste the month's data.
- 2You ask for the report in your fixed structure.
- 3You check the figures and send it.
Example prompt
Using this data, write the monthly report in this structure: three-line summary, change versus last month, issues and next steps. Don't invent data: flag anything missing.
What you gain
The report goes out on time, always in the same format.
Key figure for this department · Administration: 24 min saved writing a document · Government Digital Service (UK), 2025
The problem
Emails, lists and notes in a jumble that must be sorted by hand.
How AI solves it
- 1You paste the raw information.
- 2You ask for a classification by your categories.
- 3You review the doubtful cases.
Example prompt
Sort these requests into a table under Invoicing, Suppliers, Staff and Other. Add urgency (high, medium, low) and mark "review" on any that don't clearly fit.
What you gain
You see at a glance what to handle first.
Key figure for this department · Administration: 24 min saved writing a document · Government Digital Service (UK), 2025
The problem
Long contracts where it is easy to miss a deadline or a penalty.
How AI solves it
- 1You upload the contract.
- 2You ask for the key clauses with their reference.
- 3You validate the essentials with your adviser.
Example prompt
Summarise this contract in a table: parties, purpose, term, notice periods, amounts, penalties and renewal. Give the clause number for each point and flag what should be reviewed by a lawyer.
What you gain
No deadline slips past and you know where each clause is.
Key figure for this department · Administration: 24 min saved writing a document · Government Digital Service (UK), 2025
The problem
The same emails and documents get rewritten again and again.
How AI solves it
- 1You collect three examples you already use.
- 2You ask for a template with variable fields.
- 3You save it and share it with the team.
Example prompt
From these three payment-reminder emails, create a template with bracketed fields ([client], [amount], [date]) and three versions: first reminder, second reminder and final notice.
What you gain
The whole team uses the same text, without wasting time.
Key figure for this department · Administration: 24 min saved writing a document · Government Digital Service (UK), 2025
06 / 08
Customer support
Faster, more consistent answers, especially for less experienced staff.
In customer service, generative AI enables faster, more consistent answers, especially for less experienced staff: +15% issues resolved per hour, according to Quarterly Journal of Economics, 2025.
+15%
issues resolved per hour (5,172 agents)
+30%
among the least experienced agents
30–45%
potential productivity in customer operations
McKinsey, 2023 (estimate)
Issues resolved per hour: base 100 without AI versus 115 with AI.
Use cases
The problem
Each agent answers their own way and customers get different messages.
How AI solves it
- 1You paste the query and your policy.
- 2You ask for a clear reply based on it.
- 3You review and send it.
Example prompt
Reply to this customer using only our attached returns policy: confirm what they can do, the steps and the timeframe. Friendly tone, five lines maximum. If the policy doesn't cover the case, say so.
What you gain
Customers get the same correct answer, whoever they speak to.
Key figure for this department · Customer support: +15% issues resolved per hour (5,172 agents) · Quarterly Journal of Economics, 2025
The problem
Long email threads that must be read in full before replying.
How AI solves it
- 1You paste the history.
- 2You ask for a summary with the current status.
- 3You reply with full context.
Example prompt
Summarise this incident history in five points: initial problem, what has already been tried, what the customer expects, current status and recommended next action.
What you gain
You pick up any case without making the customer repeat themselves.
Key figure for this department · Customer support: +15% issues resolved per hour (5,172 agents) · Quarterly Journal of Economics, 2025
The problem
The same questions arrive every day and get answered from scratch.
How AI solves it
- 1You export the most frequent queries.
- 2You ask for standard replies grouped by topic.
- 3You validate and share them.
Example prompt
From these queries, identify the ten most repeated questions and write a standard reply for each, with a [personalise] field wherever the case needs adapting.
What you gain
New agents answer well from day one.
Key figure for this department · Customer support: +15% issues resolved per hour (5,172 agents) · Quarterly Journal of Economics, 2025
The problem
An angry customer and a reply that could make things worse.
How AI solves it
- 1You paste the message and your draft.
- 2You ask to adjust the tone without changing the substance.
- 3You send a reply that calms things down.
Example prompt
This customer is very upset. Rewrite my draft to acknowledge their frustration, apologise for our part and offer a concrete solution, without justifying ourselves or promising what we can't deliver.
What you gain
Difficult conversations that end better.
Key figure for this department · Customer support: +15% issues resolved per hour (5,172 agents) · Quarterly Journal of Economics, 2025
07 / 08
Operations
Tracking, incidents and reports without manual work.
In operations, generative AI lets you handle tracking, incidents and reports without manual work: −15% logistics costs, according to McKinsey (pioneering companies).
−15%
logistics costs
McKinsey (pioneering companies)
−35%
inventory
McKinsey (pioneering companies)
+65%
service level
McKinsey (pioneering companies)
Logistics costs: base 100 without AI versus 85 with AI.
Results from advanced AI planning systems. An assistant like Claude is the first step.
Use cases
The problem
Everyone does the task their own way because nothing is written down.
How AI solves it
- 1You describe the task as it is done today.
- 2You ask for a step-by-step procedure and a checklist.
- 3You test it with the team and adjust it.
Example prompt
Turn this description into a numbered procedure for goods receipt: owner, steps, quality checks and what to do if something goes wrong. Add a one-page printable checklist.
What you gain
Everyone works the same way and errors repeat less.
Key figure for this department · Operations: −15% logistics costs · McKinsey (pioneering companies)
The problem
Spreadsheets of weekly data nobody has time to interpret.
How AI solves it
- 1You upload the spreadsheet.
- 2You ask for trends, anomalies and likely causes.
- 3You check the anomalies at source.
Example prompt
Analyse this weekly order data: summarise the trend, flag anomalous weeks with the exact value, suggest possible causes and write a five-line report for management.
What you gain
You spot problems before they become emergencies.
Key figure for this department · Operations: −15% logistics costs · McKinsey (pioneering companies)
The problem
The know-how lives in two people's heads and nobody else knows how it works.
How AI solves it
- 1You record or note the explanation from whoever knows.
- 2You ask for a structured process document.
- 3The expert validates it.
Example prompt
From this transcript, document the process: objective, inputs, steps, owners, tools used and checkpoints. List any unanswered questions at the end.
What you gain
The process no longer depends on a single person.
Key figure for this department · Operations: −15% logistics costs · McKinsey (pioneering companies)
The problem
Weekly planning is done in a rush and something is always missing.
How AI solves it
- 1You list tasks, people and constraints.
- 2You ask for a balanced weekly plan.
- 3You adjust the real priorities.
Example prompt
Organise these tasks into a weekly plan for four people, respecting their availability and deadlines. Show a table per day and flag any conflicts or overload.
What you gain
An organised week where nothing gets left out.
Key figure for this department · Operations: −15% logistics costs · McKinsey (pioneering companies)
08 / 08
Education
Materials, adaptations and assessments prepared in less time.
In education, generative AI lets you prepare materials, adaptations and assessments in less time: 26 minutes saved per person per day, according to the UK Government, 2025.
26 min
saved per person per day
Use cases
The problem
Preparing a new session takes hours of research and layout.
How AI solves it
- 1You give topic, level and duration.
- 2You ask for a lesson plan with activities.
- 3You check the content and adapt it to your group.
Example prompt
Prepare a 90-minute session on this topic for adult beginners: objectives, introduction, two hands-on activities, wrap-up and materials needed. Give the timing for each part.
What you gain
You arrive with a structured, ready-to-run session.
Key figure for this department · Education: 26 min saved per person per day · UK Government, 2025
The problem
One group with very different levels and a single set of materials.
How AI solves it
- 1You paste the original text.
- 2You ask for versions for several levels.
- 3You check the key content is kept.
Example prompt
Adapt this text to three levels (beginner, intermediate and advanced) keeping the same key ideas. For each level, add three comprehension questions.
What you gain
Everyone progresses at their own pace on the same content.
Key figure for this department · Education: 26 min saved per person per day · UK Government, 2025
The problem
Creating varied exercises and their answer keys takes a lot of time.
How AI solves it
- 1You give the content covered.
- 2You ask for exercises with answers.
- 3You check the difficulty before using them.
Example prompt
Create an assessment on this content: eight multiple-choice questions with one correct answer, two open questions and an annotated answer for each. Increasing difficulty.
What you gain
Complete assessments, answer key included.
Key figure for this department · Education: 26 min saved per person per day · UK Government, 2025
The problem
Regulations and reference documents too long for learners.
How AI solves it
- 1You upload the document.
- 2You ask for a summary of the key ideas.
- 3You check it against the original.
Example prompt
Summarise this document as a study sheet: five key ideas, important terms with definitions and a final outline. Give the section of the original for each idea.
What you gain
Learners get to the essentials without losing the reference.
Key figure for this department · Education: 26 min saved per person per day · UK Government, 2025
Figures from independent studies on generative AI. They are results measured in those contexts, not a guarantee: your gain will depend on your tasks and how AI is applied in your company.
Sources
- Science study on writing productivity with AI (Noy and Zhang, 2023)
- Generative AI at Work, Quarterly Journal of Economics (Brynjolfsson, Li and Raymond, 2025)
- UK Government experiment with 20,000 civil servants (2025)
- Microsoft Work Trend Index: Copilot's earliest users (2023)
- Stanford and MIT study on AI in accounting (Choi and Xie, 2025)
- LinkedIn, Future of Recruiting 2025
- Salesforce, State of Sales statistics
- McKinsey: generative AI in marketing
- McKinsey, The economic potential of generative AI (2023) (cited without link)
- McKinsey, AI-enabled supply chain (cited without link)
Go further by role
AI training for executives and managers
AI training for executives and managers: decide faster, prepare your meetings and set the rules for AI use across your teams. Online or on-site.
See the pageAI training for marketing teams
AI training for marketing teams: write faster, adapt your content and set the rules for AI use. Online or on-site.
See the pageAI training for sales teams
AI training for sales teams: prepare meetings and write proposals and follow-ups with ChatGPT, Claude and Gemini. Online or on-site.
See the pageAI training for HR teams
AI training for HR teams: job ads, documentation and internal communication with ChatGPT, Claude and Gemini, within GDPR and EU AI Act limits.
See the pageAI training for admin and support staff
AI training for admin and support staff: emails, meeting notes, documents and organisation with ChatGPT, Claude and Gemini. No jargon, from scratch.
See the pageAI training for customer service teams
AI training for customer service teams: faster, more consistent replies with ChatGPT, Claude and Gemini. Study of 5,172 agents: +15% per hour.
See the pageFrequently asked questions
Which department should we start with?
The one with the most written, repetitive tasks: often admin, customer service or marketing. The free 30-minute diagnostic helps determine it.
Are these gains guaranteed?
No. They are results measured by studies in specific contexts. The real gain depends on each team's tasks.
Should each department be trained separately?
Both are possible. The full training brings several departments together; a role-specific training goes further into the tasks of a single team.
Would you like to know how AI can help your company?
Book a free 30-minute diagnosis: I'll work with you to identify where artificial intelligence can save your team time.