Get Started with Excel Copilot: Your First Practical Step Today
Excel Copilot: The Essentials in One Article — Real Code, Diagrams, and Concrete Steps, Excerpts from a 40-Lesson Course.
The best way to learn Excel Copilot is by doing. This article gives you a head start with practical excerpts from a 40-lesson course — enough to get your first result today.
- Introduction and Activation of Copilot
- AI-Assisted Advanced Formulas
- Pivot Tables with Copilot
- Data Cleaning and Preparation
- Power Query ETL in Excel
Conditional Formulas — Nested IF, IFS
Learning objectives
- Write a simple IF formula, then enrich it with AND and OR
- Build nested IFs for multiple tiers
- Replace nested IFs with the more readable IFS
- Use IFS.MULTIPLE to compare a value against a list of cases
- Avoid common errors: condition order, parentheses, return type
The intuition: a cascade of yes/no questions
A conditional formula asks Excel a question: "Does this value exceed 1000?". If the answer is yes, Excel returns one result; if no, another. This is exactly the reasoning you would apply manually to classify orders as "large" or "small". The IF function formalizes this reasoning.
The trap appears when there are more than two cases. Imagine assigning a grade: "Excellent", "Good", "Pass", "Fail". You need several chained questions. This is where nested IFs and their more modern alternatives come in.
IF
A single question, two outcomes. The basic building block of all conditional logic.
Nested IF
An IF inside another IF. Allows multiple tiers but quickly becomes unreadable.
IFS
The modern version: a list of condition/result pairs, without nesting.
Simple IF, then combined with AND and OR
The basic syntax is =IF(test; value_if_true; value_if_false). The test is a comparison that returns TRUE or FALSE.
Nested IFs for multiple tiers
To assign a grade based on a score, we chain IFs. Each IF handles one tier, and the "else" passes control to the next IF.
| Criterion | Nested IFs | IFS |
|---|---|---|
| Readability | Poor beyond 3 tiers | Excellent |
| Parentheses to close | One per IF | No nesting |
| "Else" case | Last argument | TRUE condition |
| Availability | All versions | Excel 2019 and later |
IFS.MULTIPLE: compare against a list of cases
IFS.MULTIPLE compares an expression to a series of exact values, like a switch. Useful for translating a code into a label.
When to avoid it
When you test ranges (score >= 90). IFS is more suitable because it handles inequalities.
LAMBDA and custom formulas
Learning objectives
- Understand the value of a reusable custom function
- Write a LAMBDA formula and test it before naming it
- Register a LAMBDA as a defined name to call it anywhere
- Use LET to name intermediate results
- Get a glimpse of MAP and REDUCE to apply logic across ranges
The intuition: wrapping a formula inside a function
Imagine you often calculate the price including tax from a price excluding tax: price * 1.2. If the logic becomes complex (variable rate, rounding, discount), you copy the same long formula everywhere. The day the rule changes, you have to fix it in dozens of cells.
LAMBDA solves this problem. You define the logic once as a function with parameters, give it a name, and then call it like any other Excel function. A single source of truth.
Basic LAMBDA syntax
A LAMBDA first lists its parameters, then the final calculation. You can test it directly by placing the arguments in parentheses at the end.
Naming a LAMBDA for reuse
Via Formulas → Name Manager → New, give it a name (e.g. PriceIncTax) and paste the LAMBDA without the test arguments. It becomes callable throughout the workbook.
Limit
Available on Microsoft 365 only. A workbook opened in an older version of Excel will not evaluate your LAMBDAs.
LET: naming intermediate calculations
Often used with LAMBDA, the LET function lets you store intermediate results under a name, making long formulas readable and faster (a reused calculation is evaluated only once).
First Copilot prompt in Excel
Learning objectives
- Open the Copilot pane and recognize its main areas
- Write a first clear prompt and obtain a result
- Understand the anatomy of a good Excel prompt
- Iterate: rephrase when the response is not perfect
- Distinguish a vague prompt from a precise one
Opening the Copilot pane
Once your three prerequisites are met (license, cloud, table), click the Copilot button in the Home tab, all the way to the right of the ribbon. A pane opens on the right side of the screen. This is your dialogue area with the AI.
Input area
At the bottom of the pane, the field where you type your request in French. Press Enter to send.
Response area
In the center, Copilot displays its response: text, formula preview, or an "Insert" button to apply it.
Suggestions
Copilot often suggests pre-filled prompts adapted to your data. Click to try them.
Our example dataset
Throughout the course we will use a simple sales table. Imagine this structured table named Sales2025:
Anatomy of a good prompt
An effective prompt generally contains four elements. The more precise your request, the better the response.
| Element | Role | Example |
|---|---|---|
| Action | The verb: calculate, create, summarize, sort | "Calculate" |
| Object | The data or column concerned | "the total sales amount" |
| Criterion | The filter or dimension | "by region" |
| Format | The expected result shape | "as a pivot table" |
Vague prompt
"Analyze my sales." → Copilot does not know what to target; the response will be generic and of little use.
Precise prompt
"Calculate the total sales amount by region for 2025 and sort from highest to lowest." → Direct and actionable response.
Iterate: the art of rephrasing
Copilot does not always give the perfect answer on the first try. That is normal. The right reflex is to rephrase by adding context rather than starting from scratch.
This article covers the most useful excerpts — the complete Excel Copilot course (11 chapters, 40 lessons, corrected exercises and final project) takes you all the way.
./access-the-full-course free course: Claude CoworkFAQ
How long does it take to learn Excel Copilot?
Are there any prerequisites?
Where to start concretely?
📬 Want to receive this type of guide every week? Subscribe for free — real code, zero fluff.