Can AI Build Your Excel Tracker? Jen’s Step-by-Step Claude + Copilot Test
Jen's order notes used to live in WhatsApp threads, a paper pad and her memory. Then the weekend orders doubled, and one Sunday night she realised she could not say whether the birthday cake order had actually made money, or who still owed her a balance.
This is Part 2 of her story. In Part 1, we followed her as she turned weekend cooking into a small business. Here we open her Excel workbook and build it step by step with Claude and Microsoft Copilot: the sheets, the formulas, the payment tracking, the dashboard and, at the end, a one-click VBA macro. We also show where our first attempts went wrong, so you can skip those mistakes.
Pick Your Starting Point: Follow the Signpost
You do not have to read this guide in order. Find the sentence that sounds most like you, then follow the sign to the right tool and the right step.
Features and plans change often, so check your own screen and Microsoft's current documentation for what your version of Excel and Copilot can do.
The Kitchen Idea: Why a Workbook Is Easier Than It Looks
Jen already runs a kitchen, so we explained Excel in kitchen terms. A good workbook is organised like a good kitchen: everything has one home, and you never keep the same thing in two places.
Why this matters: if butter goes up in price, Jen changes one cell on the pantry shelf and every recipe card, menu line and order that uses it can follow. That single idea drives the whole build.
Want to see how AI can turn a small food business's scattered order records into an organised Excel system? Watch the video for a practical look at building the tracker.
Step 1: Decide What One Order Should Trigger
We did not open an AI tool first. We asked Jen what should happen the moment she types in a new birthday order for 30 people. Her answer became our blueprint.
Customer code, event date, status
Prices appear on their own
Ingredients, packaging, delivery, preparation time
Advance now, final payment later
Sales, owed, costs, profit
She also had six questions she could never answer quickly: which dishes earn the most, who still owes money, what a 30-person order really costs, which ingredients are getting dearer, how this month compares with last, and which menu to promote again. Keep your own list of questions. It becomes your dashboard later.
Step 2: Meet Your Two AI Helpers (They Work Differently)
Jen assumed Claude and Copilot were interchangeable. They are not, and knowing the difference saved us a lot of confusion.
Our plan: Claude builds the first version, then Copilot helps Jen explore it. You do not need both. Use whichever you have, and the steps still work.
The four-part prompt we used every time
Good prompts to either tool followed the same shape. It turned out to matter more than which tool we picked.
If prompting is new to you, our piece on giving AI the right information at the right moment explains why the context part does so much work.
Our Test: One Giant Prompt vs. Design First
We set up the same request twice with Claude. Both times the goal was Jen's full tracker. The results show why Step 3 exists. These observations come from one small test, not a benchmark.
| Approach | What we asked | What happened |
|---|---|---|
| Attempt 1: giant prompt | Build me a complete Excel tracker for my food business with everything I need. | A tidy-looking workbook, but advances were treated like extra sales, and one row held every dish in an order, so costs could not be split. |
| Attempt 2: design first | Design the sheets and connections first. Do not build yet. Flag any double-counting. | The plan separated Orders from Order Items, and order value from payments received, before a single formula existed. |
Relative, illustrative view of how many structural problems we had to fix afterwards (our impression, not a score):
Attempt 1: giant prompt
Attempt 2: design first
This is the same trap we describe in our look at why confident AI answers still need checking: output that looks finished is not output that is correct.
Step 3: Ask Claude to Design the Workbook (Before Building It)
Jen opened her food-business Project in Claude and pasted this. Read the last paragraph carefully, because it is the most valuable sentence in the prompt.
I run a small home-based food business: weekend orders, festive menus and party catering. I want an Excel workbook to track ingredients, recipe costing, menu prices, customer orders, the individual dishes in each order, customer payments, monthly expenses and business performance. Design separate worksheets for each. For every worksheet give me: recommended column names, what I should type in, which fields calculate automatically, which fields connect to other sheets, and the formulas. I am a beginner, so explain in simple language. Do not create the file yet. First show me the design and point out any possible double-counting or calculation problems.
Claude's design surfaced four relationships Jen had never put into words:
Takeaway: planning first lets you correct assumptions while they are still cheap to fix.
Step 4: Build the Workbook With Claude
After Jen approved the design, she asked for the file. Notice how many rules she packed into the prompt, especially the one about sample data.
Based on the approved design, create a downloadable workbook called Jen_Food_Business_Tracker.xlsx with eight sheets: Ingredients, Recipe Costing, Menu, Orders, Order Items, Payments, Expenses and Dashboard. Use clear headings, Excel tables with filters, suitable date and currency formats, dropdown lists for order status, payment method and expense category, formulas for calculated fields, and a few sample records clearly labelled as examples. Use structured table references so formulas grow when I add rows. Do not present invented figures as real records. Explain how to download and use the file.
Whether Claude can hand you a file depends on the features available to your account. If it cannot, ask it to give you each sheet's columns and formulas, and type them in yourself. When the file arrives, check that all eight sheets exist and that sample rows are clearly marked before real customer details go in.
A quick Excel concept: what is an Excel Table?
You will see names such as tblIngredients all through this guide, so it is worth five minutes to understand them. An Excel Table is a normal range of cells that Excel has given a defined structure: column names, filters, and the ability to grow by itself when you add records.
The difference at a glance. Both grids below hold the same four ingredients. Only one of them knows it is a list.
How to turn your list into a table. Six small moves:
Highlight the ingredient data, headings included
Press the keys together
Make sure "My table has headers" is ticked
The grid changes colour and gets filter arrows
It appears in the ribbon when a table cell is selected
Type tblIngredients in the Table Name box
Table names cannot contain spaces. We start ours with tbl so they are easy to spot inside a formula.
Watch it grow. This is why Jen cares. She types a new ingredient directly under the last row, and Excel pulls it into the table:
The table stretches to include the new row and applies its formatting. If the table also has calculated columns, Excel normally fills those formulas in for the new row too. It is still worth a quick glance to confirm.
Why formulas read better. Compare the two ways of adding up the same column:
Step 5: Build the Costing Engine (Six Small Moves)
This is the part Jen found hardest to picture, so we broke it into six substeps. In each one the goal is the same: one place to keep ingredient prices, and everything else follows it. Jen buys from different stores, prices change, and the same ingredient appears in several dishes. Follow the path below, and use the colour key to tell the cells you type from the cells Excel calculates.
5.1 Create the Ingredients sheet (the price list)
Goal: one list of ingredients with their units and prices. These are sample prices, so replace them with your real supplier prices.
Do this: select the data, press Ctrl + T, and name the table tblIngredients (the Step 4 mini-lesson shows how). From now on, when a supplier changes a price, this is the only place Jen edits.
5.2 Create the Recipe Costing sheet (one row per ingredient)
Goal: record what goes into one batch of a dish. Jen's homemade chocolate cake makes 10 portions per batch, so each ingredient gets its own row. Name this table tblRecipeLines.
Quantities are in the same unit as the price list: kilograms for the dry goods and litres for milk. The ingredient names must be spelled exactly as in tblIngredients, because the next step matches on them.
5.3 Make Excel fetch the price automatically
Goal: fill the UnitCost column from the price list, so Jen never types a price twice. This is the first formula, and the first place she asked Claude for help.
Claude's formula, with each part colour-coded:
=XLOOKUP([@Ingredient], tblIngredients[Ingredient], tblIngredients[UnitCost], 0)the ingredient on this recipe row
the ingredient column of the price list
the price from that same row
show 0 instead of an error
Do this: click the first cell under UnitCost in tblRecipeLines, type the formula, and press Enter. In a table, Excel normally fills the whole column. For the Butter row, it finds Butter in the price list and brings back 5.00. The result:
5.4 Calculate the cost of each ingredient line
Goal: quantity used times price, row by row. Jen's request was one plain sentence.
=[@QtyUsed] * [@UnitCost]
Butter check: 0.5 kg × $5.00 = $2.50
5.5 Work out the cost per portion
Goal: how much ingredient cost goes into one slice of cake. Divide each line cost by the number of portions the batch makes. Jen added a safety request to her prompt, which is a good habit.
=IFERROR([@LineCost] / [@BatchYield], 0)
"Divide the line cost by the portions in the batch. If that fails, show 0."
So one portion of chocolate cake costs about $1.10 in ingredients. That is ingredients only. Packaging, labour and utilities are still to come.
5.6 Connect the recipe to the Menu sheet
Goal: put each dish's selling price right next to its ingredient cost, so Jen can see at a glance what she earns on it. Create a table named tblMenu and type the dish names and selling prices. The ingredient cost column is a formula.
=SUMIFS(tblRecipeLines[CostPerPortion], tblRecipeLines[Dish], [@Dish])"Add up the cost per portion, but only for recipe rows where the dish matches this menu row."
Only the cake has recipe lines so far, so the other two dishes show zero. That is a useful signal: a zero ingredient cost usually means a recipe has not been entered yet, not that the dish is free to make. For the cake, $1.095 against a $4.00 price means ingredients take about 27 percent of the selling price.
What Jen achieved: she has a reusable costing system instead of recalculating every dish from scratch. Here is the ripple, using the formulas above:
One cell changed in tblIngredients
0.5 kg now costs more
Batch of 10, so +0.05 per portion
Cost figures update, with no copy-pasting
Tiny per portion, but across a 25-portion order it adds up, and Jen would never have spotted it by hand.
Step 6: Build the Order Pad (Two Linked Sheets)
The mistake that sank our Attempt 1 was cramming a whole order into one row. A birthday order might have 25 main dishes, 25 desserts and a cake. The fix is two sheets joined by an Order ID, like a receipt with a header and line items.
Same colour, same order. For privacy, Jen uses customer codes rather than full names and phone numbers in the workbook.
Four formulas that make an order calculate itself
| Column in tblOrderItems | What Jen asked Claude for | Formula |
|---|---|---|
| UnitPrice | Fetch the selling price from tblMenu for the chosen dish | =XLOOKUP([@Dish],tblMenu[Dish],tblMenu[SellingPrice],0) |
| LineRevenue | Multiply Quantity by UnitPrice | =[@Quantity]*[@UnitPrice] |
| IngredientCostPerPortion | Fetch the ingredient cost per portion from tblMenu | =XLOOKUP([@Dish],tblMenu[Dish],tblMenu[IngredientCostPerPortion],0) |
| IngredientCost | Quantity times that cost per portion | =[@Quantity]*[@IngredientCostPerPortion] |
Here is what the J001 lines look like once the formulas are in. The rice and dessert cost-per-portion figures are sample values, assumed for the demonstration as if their recipes had been entered (Step 5 only covered the cake):
Then, back in tblOrders, a SUMIFS gathers every line for the order. Add an OrderRevenue column and an IngredientCost column:
OrderRevenue: =SUMIFS(tblOrderItems[LineRevenue], tblOrderItems[Order ID], [@[Order ID]])IngredientCost: =SUMIFS(tblOrderItems[IngredientCost], tblOrderItems[Order ID], [@[Order ID]])Jen typed three line items and one total appeared. The column names in your own tables must match the names in the formula exactly, spaces included. A mismatch is the most common reason a pasted formula fails.
Step 7: Separate Three Different Money Numbers
This step is where Jen had been confused for months. Suppose J001 is worth 337.50. She takes a 100.00 advance now and the remaining 237.50 after the event. How much has she sold? Not 337.50 plus 100.00 plus 237.50. She sold 337.50, once.
Payments get their own table, tblPayments, one row per payment, each tagged with an Order ID:
Here is the same order J001 over time. The order value never changes. Only the paid portion grows, and the balance shrinks to zero:
After the Nov 2 advance: paid 100.00 of 337.50 (about 30 percent)
After the Nov 8 final payment: paid 337.50 of 337.50 (100 percent)
Green = received, orange = still owed.
Two more columns go into tblOrders:
| Column | Formula | In words |
|---|---|---|
| PaidToDate | =SUMIFS(tblPayments[Amount],tblPayments[Order ID],[@[Order ID]]) | Add every payment for this order |
| BalanceDue | =MAX(0,[@OrderRevenue]-[@PaidToDate]) | Order value minus paid, never below zero |
After both J001 payments land, PaidToDate is 337.50 and BalanceDue is 0.00.
For refunds, cancellations and deposits, keep separate records and follow your customer agreement and local rules. These formulas are a simple tracker, not a full accounting system.
Step 8: Find the Real Profit of an Order
Revenue flatters you. A big catering order brings in more money but also eats more ingredients, packaging and hours. We used Part 1's 30-person birthday order, with sample figures, to show where the money goes:
Selling price: 360
Ingredients: 120
Packaging: 25
Delivery: 15
Jen's preparation time allowance: 60
Estimated contribution: 140
Bars are scaled to the selling price. To build this into the workbook, add new columns to tblOrders: four you type, then three Excel calculates.
| Column | Formula | Check with the sample |
|---|---|---|
| LaborCost | =[@LaborHours]*[@HourlyRate] | 4 × 15 = 60 |
| TotalOrderCost | =[@IngredientCost]+[@PackagingCost]+[@DeliveryCost]+[@LaborCost] | 120 + 25 + 15 + 60 = 220 |
| EstimatedContribution | =[@OrderRevenue]-[@TotalOrderCost] | 360 − 220 = 140 |
The 140 is not final profit. Overheads, tax and other costs still sit outside it, which is the subject of the next step.
Step 9: Track Monthly Expenses (And Why the Date Rule Matters)
Some spending belongs to no single order: cleaning supplies, equipment, marketing. These go in tblExpenses, kept apart from order costs so nothing is counted twice.
To total one month, the Dashboard holds a month-start date in cell B2. Jen changes that one date and every figure on the dashboard follows. She asked Claude to explain the date rule, then we drew it:
=SUMIFS(tblExpenses[Amount], tblExpenses[Date], ">="&B2, tblExpenses[Date], "<"&EDATE(B2,1))"On or after the 1st, and strictly before the 1st of next month." That second half avoids accidentally counting December 1.
With B2 set to 1 November, Jen's dashboard cell returns the sum of the four sample rows above:
Jen can total completed-order sales for a month the same way:
=SUMIFS(tblOrders[OrderRevenue], tblOrders[Status], "Completed", tblOrders[EventDate], ">="&B2, tblOrders[EventDate], "<"&EDATE(B2,1))Six measures that sound alike but are not
what customers agreed to pay
money that actually arrived
work finished in the month
money spent
what is left after order costs
needs all other costs and adjustments
Step 10: Build the Sunday-Evening Dashboard
Jen did not want to open eight sheets every Sunday. She wanted one screen. We started with six tiles, each answering one question from Step 1. Here is the layout we asked Claude to build (figures below are sample values only):
Below the tiles go three charts, each tied to a real question:
| Chart | Question it answers |
|---|---|
| Line: monthly sales trend | Is the business growing or slipping? |
| Bar: best-selling dishes | Which menu should I promote again? |
| Pie or bar: expense categories | Where is the spending going? |
Here is the prompt Jen used. Notice the rule that keeps the dashboard honest:
Using the existing tables in my workbook, create a Dashboard sheet with: completed-order sales for the selected month, payments received, outstanding balances, ingredient costs, operating expenses and estimated operating profit. Add a monthly sales trend chart, a best-selling dishes bar chart and an expense category chart. Use formulas or PivotTables linked to the existing data. Do not type independent sample figures into the dashboard; every value must be calculated from the workbook. Explain how each metric is calculated.
A good rule of thumb: if a chart does not answer a question you actually have, leave it off.
Step 11: Use Copilot Inside Excel
With a working workbook, Jen wanted to ask questions without writing formulas. This is where Copilot in Excel fits, if your Microsoft 365 plan includes it. Copilot's availability, required file location and exact controls vary by plan, organisation and Excel version, so confirm what you have before following along.
Get ready (checklist)
| Check | Why | Done? |
|---|---|---|
| Workbook open in a supported Excel version | Copilot needs a supported setup | ☐ |
| Saved to OneDrive or SharePoint, AutoSave on | Commonly required, and protects you with version history | ☐ |
| Data stored as proper Excel Tables | Gives Copilot clean column names to work with | ☐ |
| Plan includes Copilot access | Without it, skip to Step 12 | ☐ |
Three ways to work: pick the one that matches your risk
Microsoft describes different interaction modes. Treat them like a traffic light. The more it can change, the more careful you should be.
Five questions Jen asked her own data
The review habit that protects you
Whether Claude wrote it or Copilot changed it, we ran the same four checks:
Which tables and cells were touched?
Use J001: it should total 337.50
Bank statement, supplier receipts
Use undo or version history
This is the same discipline we describe in our look at when AI creates extra work: an unchecked result can cost more time than doing the task by hand.
Step 12: Make It Hard to Break
The last piece before automation is stopping messy data at the door. Three small habits did most of the work.
Jen's reusable Sunday review prompt
Review this month's business workbook. Summarise completed orders, revenue, payments collected, outstanding balances, ingredient costs, operating expenses and estimated profit. Identify the best-selling dishes, the highest ingredient-cost items, orders with unpaid balances, any unusual expense increases, and three practical improvements for next week. Keep actual figures separate from recommendations. Do not invent missing information.
Step 13: When Clicking Repeats, Ask for a Macro
Formulas calculate. They do not clean, rearrange or print. Every Monday, Jen still got a messy export, and fixing it ate an hour. That is VBA territory: a multi-step routine on desktop Excel. A macro is just a recipe card for Excel: the same steps, the same result, every time.
The messy sheet (sample rows)
| A | B | C | D | |
| 1 | OrderID | Customer | Dish | Qty |
| 2 | J001 | " Customer A " | Chocolate Cake | 25 |
| 3 | J001 | Customer A | Special Rice Dish | "25" (text) |
| 4 | J002 | (blank) | Dessert Cup | 8 |
| 5 | J002 | Customer B | Dessert Cup | 8 |
A person spots these by eye. A macro will not, unless told. Also, a duplicate may be a typing error or a genuine second order, and only Jen can tell which. So the rule we gave Claude was: flag, never delete.
The staged conversation
| Stage | Prompt | Result |
|---|---|---|
| 1. Describe | Here are sample rows from my weekly orders sheet. List what is wrong and ask what you need to know. No code yet. | Found the spaces, text-numbers and duplicates, and asked about blank customers. |
| 2. Decide | Never delete. Highlight blank customers red and possible duplicates (same OrderID and Dish) yellow. Fix spaces and text-numbers. | Rules became explicit. |
| 3. Build | Now write the VBA. Detect the last row, add error handling, comment each section. | The macro below. |
| 4. Test | Give me five test cases, including an empty sheet and a missing sheet. | The empty-sheet case exposed a gap, now handled. |
Install it in six moves
File > Save As > Excel Macro-Enabled Workbook
File > Options > Customize Ribbon > tick Developer
Press Alt + F11, then Insert > Module
Review it first, and ask AI to explain anything unclear
Click inside the code and press F5, or use Developer > Macros
Developer > Insert > Button, then assign the macro
It assumes a sheet named WeeklyOrders with headers in row 1, OrderID in column A, Customer in B, Dish in C and Qty in D. Change the names and letters if yours differ.
Sub CleanWeeklyOrders()
Dim ws As Worksheet
Dim lastRow As Long, r As Long
Dim flagged As Long
On Error GoTo ErrHandler
Set ws = ThisWorkbook.Worksheets("WeeklyOrders")
lastRow = ws.Cells(ws.Rows.Count, "A").End(xlUp).Row
If lastRow < 2 Then
MsgBox "No data rows found.", vbInformation
Exit Sub
End If
Application.ScreenUpdating = False
For r = 2 To lastRow
' 1. Remove extra spaces in Customer
ws.Cells(r, "B").Value = Trim(CStr(ws.Cells(r, "B").Value))
' 2. Convert text-numbers in Qty to real numbers
If VarType(ws.Cells(r, "D").Value) = vbString Then
If IsNumeric(Trim(ws.Cells(r, "D").Value)) Then
ws.Cells(r, "D").Value = CDbl(Trim(ws.Cells(r, "D").Value))
End If
End If
' 3. Flag blank customers in red (never delete)
If Len(ws.Cells(r, "B").Value) = 0 Then
ws.Cells(r, "B").Interior.Color = RGB(255, 199, 206)
flagged = flagged + 1
End If
' 4. Flag possible duplicates (same OrderID and Dish) in yellow
If Application.WorksheetFunction.CountIfs( _
ws.Range("A2:A" & lastRow), ws.Cells(r, "A").Value, _
ws.Range("C2:C" & lastRow), ws.Cells(r, "C").Value) > 1 Then
ws.Cells(r, "C").Interior.Color = RGB(255, 235, 156)
flagged = flagged + 1
End If
Next r
MsgBox "Done. Rows checked: " & (lastRow - 1) & ". Flags raised: " & flagged, vbInformation
CleanUp:
Application.ScreenUpdating = True
Exit Sub
ErrHandler:
MsgBox "Cleanup stopped: " & Err.Description, vbExclamation
Resume CleanUp
End Sub
After one click, spaces are gone, both text-quantities are real numbers, and the blank customer and the possible duplicate are highlighted. Nothing is removed. Jen checks with the customer, then deletes the extra row herself if it really was a double entry.
Test it before you trust it
| Test case | Expected | Pass? |
|---|---|---|
| Normal week | Flags only where expected | ☐ |
| Header row only | Friendly message, no changes | ☐ |
| Sheet renamed or missing | Clear error message, no crash | ☐ |
| 500 rows instead of 5 | All rows checked in reasonable time | ☐ |
| Run it twice | Same result, no damage | ☐ |
Simplest tool first
| Tool | Good for | Watch out for |
|---|---|---|
| Formulas and Tables | Calculations that update themselves | Very long nested formulas |
| Power Query | Repeatable import and cleanup | Initial learning curve |
| PivotTable | Summaries and grouping | Needs a refresh after data changes |
| VBA macro | Multi-step desktop routines and buttons | Desktop only, security and upkeep |
| Copilot in Excel | Plain-language questions about your data | Plan-dependent, results need checking |
For a wider look at which Excel jobs AI handles well, see our guide to Excel problems AI can solve in minutes.
15 Jobs AI Can Do for Your Macros (Quick Menu)
Cleanup was one use. Here are fifteen, grouped by what you are trying to do. Each is a good first prompt.
If we had to start with one, it would be the last: ask "what is the simplest safe way to automate this?" before "write me a macro." It is also a handy partner to our guide to automating repetitive work without losing control.
Mistakes We Made (So You Can Skip Them)
Limitations: When Not to Use This
- Looks right is not right. AI can write a formula that points at the wrong column or the wrong rows. Always test with a small example where you already know the answer.
- Not an accounting system. Refunds, deposits, taxes and cancellations need proper records and local rules. Our workbook is a tracker.
- Desktop only for VBA. Excel for the web cannot create, run or edit macros. Browser-based teams may do better with Office Scripts or Power Query, subject to plan availability.
- Plan-dependent features. File creation in Claude and Copilot in Excel depend on your account and subscription. Check your own screen.
- Privacy. Customer details do not belong in an AI chat by default. Use codes or sample data, and follow your organisation's data policy.
- Our macro has limits. The duplicate check rescans the column for every row, which is fine for hundreds of rows but slow for very large sheets, and it does not fix mixed date formats. Those would be our next prompts.
Your First Seven Days
Do not try to enter every business record on day one. Start small and prove the formulas.
List what you need to track and your six questions
Ask Claude for the design, then the workbook
Enter real ingredient prices and one recipe
Add orders and order items
Add payments and expenses
Build the dashboard and try Copilot
Test against figures you know by hand
The same blueprint adapts well beyond food. A tutor can swap dishes for sessions, a photographer can swap them for packages, and a freelance designer can swap them for projects. The structure (list, line items, payments, expenses, dashboard) stays the same. If you are choosing which Claude model to use for planning and formulas, our model tiers guide helps match the model to the task.
Final Thoughts
Jen never became an Excel expert. She learned to explain what her business needed, ask AI to plan before building, and check every result against something she already knew. Claude helped her design and build. Copilot, where available, helped her explore. A macro took care of the Monday cleanup.
The part AI cannot do for you is decide what counts as a real order, what counts as profit, and which numbers to trust. Get those rules right, and the formulas, charts and macros become surprisingly easy. If you have not yet read how Jen's business began, start with Part 1, then come back and build your own tracker.
Frequently Asked Questions
1. Can Claude create an Excel workbook with formulas and charts?
Often, yes, depending on the features available to your account. Ask for the workbook in stages and test the formulas before using real data. If file creation is not available, ask for column lists and formulas and enter them yourself.
2. Is Copilot better than Claude for Excel?
They suit different jobs. Claude is handy for designing, explaining and generating. Copilot works inside a supported workbook. The better choice depends on your plan and task.
3. Do I need to know Excel formulas first?
No. Describe the calculation in plain words and ask for the formula and an explanation. You should still understand what it does and check the result.
4. Why use Excel Tables instead of normal ranges?
Tables have names, filters and automatic expansion, and structured references such as tblOrders[OrderRevenue] are easier to read than cell ranges.
5. How do I avoid counting an advance twice?
Keep order value and payments in separate tables, linked by Order ID. Total the order once and add up payments separately.
6. Can I use Copilot in Excel without a paid plan?
Availability depends on your Microsoft 365 subscription, organisation and Excel version. Check Microsoft's current documentation and your account before following the Copilot steps.
7. Is AI-generated VBA safe to run?
Treat it like code from a stranger. Read it, ask AI to explain it, test it on a copy, and be wary of anything that touches files, networks or sensitive data.
8. Is it safe to upload my business workbook to an AI tool?
Review the provider's data policy and your organisation's rules first. Remove customer names, phone numbers, addresses and payment details, and never share passwords or financial credentials.
Comments
Post a Comment