AI for Data Analysis & Excel: A Practical Guide

If you spend a significant portion of your working week wrestling with spreadsheets, writing formulas, cleaning data, or building reports, ai data analytics tools are about to become your most valuable productivity asset. This guide is written specifically for professionals in Irish organisations who work with data in Excel, Google Sheets, or Power BI and want to understand how excel ai capabilities and tools like ChatGPT can transform their workflow — without needing a data science degree.

You'll learn how to use AI to write complex Excel formulas in seconds, clean messy data sets, create visualisations, generate reports, and extract insights from data that would have taken hours to analyse manually. We'll cover specific tools, step-by-step tutorials, and real-world examples relevant to Irish business contexts — from Revenue returns to CSO data to monthly management reporting.

If you're completely new to AI tools, consider starting with our free AI course to build your foundation before diving into data-specific applications.

AI Data Analytics Tools You Should Know About

The landscape of ai for data analysis has expanded rapidly. Here are the key tools available today, ranked by accessibility for people who primarily work in spreadsheets.

Microsoft Copilot in Excel

Microsoft Copilot is Microsoft's AI assistant, integrated directly into Excel as part of Microsoft 365. For anyone already working in Excel daily, this is the most seamless AI data tool available.

What Copilot can do in Excel:

  • Analyse data with natural language: Ask questions like "What was the highest-selling product in Q3?" and Copilot will analyse your spreadsheet and provide the answer.
  • Generate formulas: Describe what you need and Copilot will write the formula. "Calculate the year-over-year percentage change in revenue for each region."
  • Create charts: "Create a bar chart showing monthly sales by product category" — Copilot generates the chart directly in your spreadsheet.
  • Identify trends and outliers: "Highlight any months where expenses exceeded budget by more than 10%."
  • Sort, filter, and format: "Sort the data by revenue descending and highlight the top 10 rows in green."

Requirements: You need a Microsoft 365 Business Standard or Enterprise licence plus the Copilot add-on (approximately -30/user/month). Your data must be in an Excel Table format (not just a range) and stored in OneDrive or SharePoint.

ChatGPT Advanced Data Analysis

ChatGPT's Advanced Data Analysis feature (available in the paid tier at -20/month) allows you to upload Excel files, CSV files, and other data formats directly into the conversation. ChatGPT can then analyse the data, create visualisations, perform statistical analysis, and answer questions — all within the chat interface.

This is particularly powerful because:

  • You don't need to know any programming language — you just describe what you want in plain English.
  • ChatGPT writes and executes Python code behind the scenes to analyse your data.
  • It can handle complex analyses that would be difficult or impossible in Excel alone.
  • You can download the results, including cleaned data sets, charts, and modified spreadsheets.

Google Sheets AI Features

Google has integrated AI capabilities into Google Sheets through Gemini:

  • Help me organise: Describe the spreadsheet you need, and Gemini will create a structured template with headers, sample data, and formatting.
  • Formula suggestions: As you type formulas, Google Sheets offers AI-powered suggestions and corrections.
  • Smart Fill: Automatically detects patterns in your data and fills cells accordingly — similar to Flash Fill in Excel but powered by AI.
  • Natural language queries: In Google Sheets, you can type questions about your data in the Explore panel and receive answers with supporting charts.

Comparison Table: AI Data Tools

Tool Best For Cost Data Privacy Skill Level
Copilot in Excel Daily Excel users, integrated workflow ~-30/user/month (plus M365) Enterprise-grade (data stays in M365) Beginner
ChatGPT Advanced Data Analysis Complex analysis, visualisation, ad-hoc queries -20/month Good (data deleted after session, but check terms) Beginner-Intermediate
Google Sheets + Gemini Google Workspace users, collaboration Included with Workspace / -19.99 for Gemini Advanced Enterprise-grade for Workspace accounts Beginner
Claude Large document analysis, careful reasoning -20/month Good (does not train on your data) Beginner-Intermediate
Power BI Copilot Business intelligence, dashboards, reporting Included with Power BI Premium Enterprise-grade Intermediate

Using ChatGPT to Write Excel Formulas: Step-by-Step with Examples

This is one of the most immediately valuable applications of AI for anyone who works with spreadsheets. Instead of spending 20 minutes searching for the right formula on Google, you can describe what you need in plain English and get a working formula in seconds.

How to Ask ChatGPT for Excel Formulas

The key to getting accurate formulas is to be specific about your data layout. Always tell ChatGPT:

  1. What columns contain what data (e.g., "Column A has customer names, Column B has dates, Column C has amounts")
  2. Where your data starts and ends (e.g., "Data runs from row 2 to row 500, with headers in row 1")
  3. Exactly what result you want
  4. Any conditions or exceptions

Example 1: VLOOKUP / XLOOKUP

Prompt: "I have two Excel sheets. Sheet1 has a list of invoice numbers in column A and customer names in column B. Sheet2 has invoice numbers in column A and payment amounts in column C. I need a formula in Sheet1, column C, that looks up each invoice number and returns the payment amount from Sheet2. Some invoices may not have a payment yet — in that case, show 'Unpaid'."

ChatGPT's response will include:

For modern Excel (Office 365):

=XLOOKUP(A2,Sheet2!A:A,Sheet2!C:C,"Unpaid")

For older Excel versions:

=IFERROR(VLOOKUP(A2,Sheet2!A:C,3,FALSE),"Unpaid")

ChatGPT will also explain what each part of the formula does, which helps you learn as you go.

Example 2: SUMIFS with Multiple Criteria

Prompt: "My spreadsheet has sales data. Column A is the salesperson name, Column B is the region (Dublin, Cork, Galway, Limerick), Column C is the date, Column D is the sale amount. I need a formula to calculate total sales for salesperson 'Sarah Murphy' in the 'Dublin' region for dates in Q1 2026 (January to March 2026)."

Result:

=SUMIFS(D:D,A:A,"Sarah Murphy",B:B,"Dublin",C:C,">="&DATE(2026,1,1),C:C,"<="&DATE(2026,3,31))

This is a formula that many intermediate Excel users would struggle to write correctly from memory. With AI, you get it in seconds.

Example 3: Conditional Formatting Formula

Prompt: "I want to use conditional formatting in Excel. My data is in columns A to F, rows 2 to 100. Column E contains 'Status' with values like 'Overdue', 'On Track', 'Complete', and 'At Risk'. I want: Overdue rows highlighted in light red, At Risk rows in light yellow, Complete rows in light green, and On Track rows with no highlighting. Give me the conditional formatting formulas I need."

ChatGPT will provide:

  • For red (Overdue): =$E2="Overdue"
  • For yellow (At Risk): =$E2="At Risk"
  • For green (Complete): =$E2="Complete"

Along with step-by-step instructions for applying these via the Conditional Formatting Rules Manager.

Example 4: Pivot Table Guidance

While ChatGPT can't create a pivot table for you directly (that requires interaction with Excel), it can tell you exactly how to set one up:

Prompt: "I have a data set with columns: Date, Department, Expense Category, Amount, Approved By. I need a pivot table that shows total expenses by department, broken down by expense category, with monthly columns. Walk me through setting this up step by step in Excel."

ChatGPT will provide detailed, numbered steps including which fields to drag to which areas (Rows, Columns, Values, Filters) of the PivotTable Fields panel.

Example 5: Array Formulas and Dynamic Arrays

Prompt: "I need a formula that returns a sorted list of unique customer names from column A (which has many duplicates), sorted alphabetically. I'm using Excel 365."

Result:

=SORT(UNIQUE(A2:A1000))

For older Excel versions, ChatGPT will provide the more complex legacy alternative using INDEX, MATCH, and COUNTIF combinations.

Using AI to Clean Messy Data

Anyone who works with real-world data knows that a significant portion of the job is cleaning it. Inconsistent formatting, duplicate entries, missing values, merged cells, mixed data types — these issues plague every data set. AI can dramatically speed up the cleaning process.

Common Data Cleaning Tasks AI Can Handle

Standardising Names and Addresses

Irish business data is notoriously inconsistent when it comes to names and addresses. The same company might appear as "O'Brien & Sons Ltd", "O'Brien and Sons Limited", "OBrien & Sons", and "O'Brien & Sons Ltd." in different records.

Prompt for ChatGPT: "I'm going to paste a list of company names from my spreadsheet. Many are duplicates with slight variations in spelling, punctuation, or formatting. Please identify the duplicates and suggest a standardised version for each group."

For larger data sets, upload the file to ChatGPT Advanced Data Analysis:

Prompt: "This spreadsheet contains customer records. Column B has company names with many inconsistencies and duplicates. Please identify groups of records that appear to be the same company, standardise the names, and create a new column with the cleaned names."

Fixing Date Formats

Date formats are a perennial headache, especially in Ireland where we use DD/MM/YYYY but imported data often uses the American MM/DD/YYYY format.

Prompt: "My Excel column has dates in mixed formats — some are DD/MM/YYYY, some are MM/DD/YYYY, some are written as '15 March 2026', and some are just '15/3/26'. I need a formula or approach to standardise all dates to DD/MM/YYYY format. The data is in column C, rows 2 to 500."

ChatGPT will provide a combination of formulas and step-by-step instructions for handling each format variant.

Removing Duplicates Intelligently

Excel's built-in Remove Duplicates feature is blunt — it only catches exact matches. AI can handle fuzzy matching:

Upload your file to ChatGPT and prompt: "This spreadsheet has a customer list with duplicates. Some duplicates are exact matches, but others have slight variations (typos, different abbreviations, extra spaces). Please identify all likely duplicates, group them, and create a clean list keeping the most complete record from each group."

Splitting and Combining Data

A common Irish data issue: full addresses in a single cell that need to be split into Address Line 1, Address Line 2, Town, County, and Eircode.

Prompt: "Column A contains full Irish addresses in a single cell, like '15 Main Street, Rathmines, Dublin 6, D06 X2Y3'. I need formulas to split these into: Street Address, Area/Town, County/City, and Eircode in separate columns."

ChatGPT will provide formulas using TEXT functions or suggest using Flash Fill / Power Query approaches depending on data consistency.

Step-by-Step: Cleaning a Data Set with ChatGPT Advanced Data Analysis

Here's a complete walkthrough for cleaning a typical Irish business data set:

  1. Open ChatGPT (paid version) and start a new conversation.
  2. Upload your file by clicking the paperclip icon and selecting your Excel or CSV file.
  3. Ask for an overview: "Please examine this data set and give me a summary: number of rows and columns, data types, any obvious quality issues (missing values, duplicates, inconsistent formatting)."
  4. Review the assessment — ChatGPT will typically identify issues you hadn't noticed.
  5. Request cleaning: "Please clean this data set by: (1) removing exact duplicate rows, (2) standardising the date format to DD/MM/YYYY, (3) trimming extra spaces from all text fields, (4) filling in missing county values where the Eircode is present, (5) standardising company names to title case."
  6. Download the result: ChatGPT will provide a download link for the cleaned file.
  7. Verify: Always spot-check the cleaned data. Open it in Excel and review a sample of records to ensure the cleaning was done correctly.

AI for Data Visualisation

Creating effective charts and visualisations is both an art and a science. AI tools can help with both — suggesting the right chart type for your data and actually creating the visualisation.

Using ChatGPT for Visualisation

When you upload data to ChatGPT Advanced Data Analysis, you can request visualisations directly:

Prompt: "Create a visualisation showing monthly revenue trends for 2025 vs 2026, with each year as a separate line. Use a professional colour scheme. Add a title and axis labels."

ChatGPT will generate a chart using Python's matplotlib or seaborn libraries. You can then refine it:

  • "Make the lines thicker and add data point markers."
  • "Change the colour scheme to our brand colours (navy #003366 and green #009933)."
  • "Add a horizontal line showing the annual target of -500,000."
  • "Export this as a high-resolution PNG suitable for a presentation."

Choosing the Right Chart Type

Not sure which chart to use? Ask AI:

Prompt: "I have data showing quarterly sales broken down by five product categories across four regions over two years. I need to present this to the board and highlight: (1) overall growth trend, (2) which regions are performing best, and (3) which product categories are growing fastest. What chart types would you recommend and why?"

ChatGPT will recommend specific chart types for each insight — perhaps a line chart for trends, a grouped bar chart for regional comparison, and a stacked area chart for category growth — and explain the reasoning behind each recommendation.

Excel Chart Enhancement

If you prefer to create charts in Excel but want them to look more professional:

Prompt: "I've created a basic bar chart in Excel showing monthly expenses by department. Give me step-by-step instructions to make it look professional: better colours, cleaner layout, proper labels, no chart junk. I want it suitable for a board report."

ChatGPT will provide detailed formatting instructions including specific colour codes, font sizes, gridline settings, and layout tips that transform a default Excel chart into something presentation-ready.

AI for Report Writing from Data

One of the most time-consuming tasks for anyone working with data is translating numbers into narrative — writing the monthly report, the quarterly review, the management commentary. AI can dramatically accelerate this process.

Generating Report Narratives from Data

Here's a practical workflow:

  1. Upload your data to ChatGPT (or paste the key figures if the data is sensitive).
  2. Request analysis: "Analyse this data and identify the five most important trends, insights, or anomalies that management should be aware of."
  3. Generate the narrative: "Write a management report section based on this analysis. Use a professional but accessible tone. Include specific figures. Structure it with an executive summary, key findings, areas of concern, and recommendations."
  4. Refine: "Add more detail on the revenue decline in the Connacht region" or "Include a comparison with the same quarter last year."
  5. Review and edit: Always add your own expertise, context, and judgement to the AI-generated narrative. The AI doesn't know about the new competitor that opened in Galway or the supply chain issue that affected March figures.

Monthly Financial Reporting Example

Prompt: "Here are our monthly P&L figures for the past 12 months [paste data]. Write a financial commentary for the board covering: revenue performance vs budget, gross margin trends, operating cost highlights, and a forward outlook. Use Euro figures. Keep it to approximately 500 words."

The output will be a structured, professional narrative that you can then refine with your knowledge of the business context. What might have taken 90 minutes to write from scratch now takes 15 minutes to review and customise.

Automated Insights for Regular Reports

For reports you produce regularly (monthly, quarterly), create a reusable prompt template:

Template prompt: "You are a financial analyst writing a monthly performance report for an Irish SME. I will provide the latest month's data. Please: (1) Compare to the previous month and same month last year, (2) Identify the three most significant changes, (3) Flag any figures that are more than 15% above or below budget, (4) Write a 300-word executive summary. All figures in Euro. Use British/Irish English."

Save this prompt and reuse it each month, simply updating the data. Your report writing time drops from hours to minutes.

Power BI Copilot: AI-Powered Business Intelligence

For organisations using Microsoft Power BI for dashboards and business intelligence, Copilot adds a powerful AI layer. Our Microsoft Copilot guide covers this in detail, but here are the essentials for data professionals.

What Power BI Copilot Can Do

  • Create reports from natural language: "Create a report page showing sales performance by region for the last 12 months, with a comparison to targets."
  • Generate DAX formulas: DAX (Data Analysis Expressions) is Power BI's formula language and can be complex. Copilot can generate DAX measures from plain English descriptions.
  • Summarise reports: Click the Copilot button on any report page and ask "Summarise the key insights from this dashboard." Copilot will analyse the visuals and data to provide a written narrative.
  • Answer questions: "Why did revenue drop in February?" — Copilot will analyse the underlying data to identify contributing factors.
  • Suggest report pages: Based on your data model, Copilot can suggest additional report pages and visualisations you might find useful.

Getting Started with Power BI Copilot

Requirements include a Power BI Premium or Fabric capacity licence and data stored in a supported format. For Irish SMEs, this may be a significant investment, so consider whether your reporting needs justify the cost or whether ChatGPT Advanced Data Analysis can meet your needs at a lower price point.

Step-by-Step Tutorial: Analysing a Spreadsheet with ChatGPT

Let's walk through a complete, realistic scenario from start to finish. Imagine you're an operations manager at an Irish company and you've been given a spreadsheet of the past year's customer support tickets.

Step 1: Upload and Explore

Open ChatGPT (paid version), click the paperclip icon, and upload your file. Then type:

"I've uploaded our customer support ticket data for 2025. Please examine the file and tell me: (1) How many records are there? (2) What columns are included? (3) What date range does it cover? (4) Are there any data quality issues I should know about?"

ChatGPT will read the file and provide a comprehensive overview. You might learn that there are 4,327 records, spanning January to December 2025, with columns for Ticket ID, Date, Customer, Category, Priority, Assigned To, Resolution Time (hours), and Status. It might also flag that 43 records have missing resolution times and 12 have invalid date entries.

Step 2: Ask Questions

Now start asking the questions you'd normally spend hours answering manually:

"What are the top five most common ticket categories, and what percentage of total tickets does each represent?"

"What's the average resolution time by category? Which category takes the longest to resolve?"

"Show me the monthly trend in ticket volume. Is it increasing, decreasing, or stable?"

"Which team member has the fastest average resolution time? Which has the slowest?"

"Are there any patterns in when tickets are submitted? Day of week? Time of day?"

Each question will get an immediate, data-backed answer — often with a visualisation included.

Step 3: Deep Dive

Based on the initial answers, go deeper:

"You mentioned that 'Billing' tickets have the longest resolution time. Break this down further — is the long resolution time consistent, or are there a few extreme outliers skewing the average? Show me a box plot."

"Is there a correlation between ticket priority and resolution time? I'd expect high-priority tickets to be resolved faster."

"Run a month-over-month comparison of ticket volume and average resolution time. Are we getting faster or slower at resolving issues?"

Step 4: Generate a Report

"Based on all the analysis we've done, write a management report titled 'Customer Support Performance Review — 2025'. Include: executive summary, key metrics, trend analysis, areas of concern, team performance summary, and three specific recommendations for improvement. Format it professionally with sections and bullet points."

Step 5: Export and Verify

Ask ChatGPT to export the cleaned data, charts, and report. Then — and this is crucial — verify the key figures. Spot-check a few numbers against the original data. AI is powerful but not infallible, and a single incorrect figure in a management report can undermine your credibility.

Limitations and Pitfalls: What AI Gets Wrong with Data

AI is a powerful tool for data analysis, but it has important limitations that every user must understand. Ignoring these can lead to serious errors in your analysis and reporting.

Hallucinated Calculations

When using text-based AI (like ChatGPT without the Advanced Data Analysis feature), be aware that the AI may "hallucinate" numbers. If you paste data into a regular ChatGPT conversation and ask for calculations, the AI might generate plausible-looking but incorrect figures because it's predicting what the answer should look like rather than actually computing it.

Rule of thumb: Always use the Advanced Data Analysis feature (which runs actual Python code) for numerical analysis. Never rely on calculations from a standard text-based chat.

Context Misinterpretation

AI doesn't understand the business context behind your data. It doesn't know that:

  • The revenue dip in March was because of a bank holiday and one fewer working day.
  • "N/A" in the status column means "Not Applicable," not "missing data."
  • The Dublin region includes Wicklow for your company's reporting purposes.
  • Figures before July 2025 are in a different format because you changed accounting systems.

Always provide context with your data, and always review AI analysis with your domain knowledge.

Sampling and Size Limitations

Different AI tools have different limits on how much data they can process:

  • ChatGPT Advanced Data Analysis can handle files up to approximately 512 MB, but performance degrades with very large files. For data sets over 100,000 rows, consider summarising or sampling first.
  • Copilot in Excel works best with Excel Tables under 2 million cells.
  • Pasting data into a chat window is limited by the AI's context window. For any serious analysis, upload the file rather than pasting.

Privacy and Confidentiality

This is especially important for Irish organisations subject to GDPR:

  • Never upload files containing personal data (customer names, addresses, PPS numbers, financial details) to consumer-grade AI tools without appropriate safeguards.
  • Anonymise or pseudonymise data before uploading. Replace customer names with Customer001, Customer002, etc. Remove identifying details.
  • Use enterprise-tier tools (ChatGPT Team/Enterprise, Copilot for Microsoft 365) when working with business-sensitive data, as these have stronger data protection commitments.
  • Check your organisation's AI use policy before uploading any business data to an AI tool.

Verification is Non-Negotiable

We cannot stress this enough: always verify AI-generated analysis. A practical approach:

  1. Check totals against known figures (e.g., cross-reference revenue totals with your accounting system).
  2. Spot-check individual calculations on a sample of records.
  3. Verify that filters and criteria were applied correctly.
  4. Have a colleague review key findings before they go to management.

Irish Business Data Context: Practical Applications

Working with data in an Irish business context comes with specific requirements and opportunities. Here's how AI can help with common Irish data tasks.

Revenue Returns and Tax Compliance

Irish businesses dealing with Revenue returns (VAT3, PAYE, corporation tax) often need to extract and format data from their accounting systems. AI can help:

  • VAT reconciliation: Upload your sales and purchase ledgers and ask ChatGPT to reconcile VAT figures, identifying any discrepancies between your records and what's been filed.
  • Expense categorisation: For Revenue compliance, expenses need to be properly categorised. AI can help classify transactions: "Review this list of transactions and categorise each as: Motor Expenses, Travel & Subsistence, Professional Fees, Office Supplies, Utilities, or Other. Flag any that seem unusual or might need manual review."
  • Year-end preparation: Ask AI to prepare a summary of key financial figures needed for your annual return, formatted to match Revenue requirements.

Important caveat: Always have your accountant or tax advisor review any Revenue-related figures. AI can speed up preparation, but tax compliance requires professional oversight.

CSO Data Analysis

The Central Statistics Office (CSO) publishes a wealth of data that Irish businesses can use for market analysis, benchmarking, and strategic planning. AI can help you make sense of it:

Prompt: "I've downloaded the CSO data on retail sales by sector for the past five years. Analyse this data and tell me: which retail sectors are growing fastest, which are declining, and how does the overall trend compare to inflation? I'm particularly interested in the implications for a mid-size retailer in the food and beverage sector."

This type of analysis — combining multiple data points with business context — would take hours manually but can be completed in minutes with AI.

Business Reporting for Irish SMEs

Irish SMEs typically need to produce regular reports for various stakeholders:

  • Management accounts: Monthly or quarterly financial summaries for directors and management.
  • Enterprise Ireland reporting: If you've received Enterprise Ireland funding, you'll need to report on specific KPIs. AI can help structure these reports and ensure all required metrics are included.
  • Bank reporting: Businesses with bank facilities often need to provide regular financial updates. AI can help generate professional reports that present your figures in the best light while remaining accurate.
  • Grant applications: Many Irish grants (Enterprise Ireland, LEO, -darás na Gaeltachta) require detailed financial projections. AI can help model different scenarios and present the data compellingly.

Eircode and Geographic Data

Irish geographic data has its own quirks. AI can help with:

  • Eircode validation: "Check this list of Eircodes and flag any that don't match the standard format (A65 F4E2 pattern)."
  • Regional analysis: "Categorise these customer records by province (Leinster, Munster, Connacht, Ulster) based on their county field, and create a summary showing customer count and revenue by province."
  • Distance calculations: For delivery planning or territory mapping, AI can help calculate distances between Eircodes or towns.

Advanced Techniques: Getting More from AI Data Analysis

Once you're comfortable with the basics, these advanced techniques will help you extract even more value from AI data tools.

Creating Reusable Analysis Templates

For reports you run regularly, create a ChatGPT "Custom GPT" or save a detailed prompt template:

Template: "You are a data analyst for an Irish SME. When I upload a monthly sales file, always perform the following analysis: (1) Total revenue vs previous month and same month last year, (2) Revenue by product category with percentage change, (3) Top 10 customers by revenue, (4) Average order value trend, (5) Regional breakdown (Dublin, Rest of Leinster, Munster, Connacht/Ulster), (6) Flag any customer whose spending dropped more than 20% vs previous month, (7) Create three charts: revenue trend, category breakdown, and regional heat. Output a structured report with all figures in Euro, using Irish English."

Combining Multiple Data Sources

AI can merge and analyse data from different sources — something that's painful to do manually in Excel:

"I'm uploading three files: (1) our sales data for Q4 2025, (2) our customer satisfaction survey results, and (3) our customer support ticket data. Please merge these on customer ID and analyse: Is there a relationship between support ticket volume and customer satisfaction? Do our highest-spending customers have more or fewer support issues? Are there customers with high spending but low satisfaction who might be at risk of churning?"

Predictive Analysis

ChatGPT Advanced Data Analysis can perform basic predictive modelling:

"Based on our monthly revenue data for the past three years, create a forecast for the next six months. Use an appropriate time-series method. Show the forecast with confidence intervals and explain the methodology."

While this won't replace a professional statistical analysis, it gives you a reasonable starting point for budgeting and planning.

Scenario Modelling

AI excels at "what if" analysis:

"Our current pricing is -45 per unit with an average monthly volume of 2,000 units. Model the following scenarios: (1) 10% price increase with estimated 5% volume drop, (2) 5% price decrease with estimated 12% volume increase, (3) introducing a premium tier at -65 with estimated 15% of current customers upgrading. For each scenario, show the impact on monthly revenue, gross margin (current margin is 35%), and annual revenue. Present in a comparison table."

Excel Formulas Cheat Sheet: Common Requests and AI-Generated Solutions

Here's a quick reference table of common Excel tasks and the prompts to use with ChatGPT to get instant formulas:

Task Prompt to Use Typical Formula Result
Look up a value "Look up [value] in [range] and return [column]" =XLOOKUP() or =VLOOKUP()
Sum with conditions "Sum [column] where [criteria column] equals [value]" =SUMIF() or =SUMIFS()
Count with conditions "Count rows where [column] contains [text]" =COUNTIF() or =COUNTIFS()
Calculate percentage change "Percentage change between [old value] and [new value]" =(New-Old)/Old
Remove duplicates with formula "Unique values from [range], sorted" =SORT(UNIQUE())
Concatenate with conditions "Join text from [column] where [criteria]" =TEXTJOIN() with IF()
Extract text "Extract text before the first space in [cell]" =LEFT(A1,FIND(" ",A1)-1)
Working days between dates "Business days between [date1] and [date2], excluding Irish bank holidays" =NETWORKDAYS() with holiday list
Running total "Running cumulative sum of [column]" =SUM($D$2:D2)
Rank values "Rank [values] from highest to lowest" =RANK() or =RANK.EQ()

Building an AI-Enhanced Data Workflow

Rather than using AI tools in isolation, the real productivity gains come from integrating them into a coherent workflow. Here's a recommended approach for Irish professionals who work with data regularly:

Daily Data Workflow

  1. Data collection: Use your existing systems (accounting software, CRM, ERP) to extract raw data. Export to Excel or CSV.
  2. Data cleaning: Upload to ChatGPT Advanced Data Analysis for cleaning if needed, or use Copilot in Excel for smaller fixes.
  3. Analysis: Use Copilot in Excel for quick, in-spreadsheet analysis. Use ChatGPT for deeper analysis, cross-referencing, or complex questions.
  4. Visualisation: Create charts in Excel with Copilot assistance, or use ChatGPT to generate more sophisticated visualisations.
  5. Reporting: Use AI to draft the narrative section of your report, combining data insights with your business context.
  6. Review: Always verify key figures manually. Have a colleague cross-check important reports.

Monthly Reporting Workflow

  1. Week 1: Collect and clean data from all sources.
  2. Week 2: Run standard analyses using saved prompt templates.
  3. Week 3: Generate report drafts with AI, review and refine with human expertise.
  4. Week 4: Finalise reports, present to stakeholders, and capture feedback for next month's improvements.

This structured approach ensures consistency while taking full advantage of AI efficiency.

Further Learning and Related Resources

Data analysis with AI is a rapidly evolving field. Here are resources to continue developing your skills:

Conclusion: From Spreadsheet Struggling to Data Confidence

The gap between "good with data" and "struggling with spreadsheets" has never been smaller. AI tools — particularly ChatGPT Advanced Data Analysis and Microsoft Copilot in Excel — have democratised data analysis in a way that was unimaginable even two years ago. You no longer need to memorise complex formulas, learn programming languages, or spend hours manually cleaning data.

What you do need is the confidence to start, the knowledge of what's possible, and the good judgement to verify AI outputs before acting on them. This guide has given you the practical steps and real-world examples to begin transforming how you work with data.

The professionals who thrive in the coming years won't be those who can write the most complex VLOOKUP. They'll be the ones who can ask the right questions of their data, use AI tools to get answers quickly, and combine those answers with business expertise to make better decisions.

Start with our free 2-hour AI Essentials course to build your foundation, and then apply what you've learned in this guide to your very next spreadsheet. Your data workflow will never be the same.