Skip to content
ClaudeClaude

Getting started with Claude in Excel

Claude understands your entire workbook—from nested formulas to multiple tab dependencies. Get explanations with cell-level citations, and update assumptions while preserving formulas. Now in beta as a research preview for all Claude Pro, Max, Team and Enterprise plan customers. Learn more: https://claude.com/claude-in-excel

Jan 30, 20267mWatch on YouTube ↗

CHAPTERS

  1. 0:03 – 0:34

    What Claude in Excel is and how to open it (keyboard shortcuts)

    The video introduces Claude in Excel as an AI agent embedded directly in spreadsheets to help with repetitive and analytical work across common business functions. It also shows the keyboard shortcuts to open Claude’s sidebar on Mac and Windows.

    • Claude works inside Excel to assist with manual spreadsheet tasks
    • Useful across finance, accounting, and operations workflows
    • Open Claude with Ctrl+Option+C (Mac) or Ctrl+Alt+C (Windows)
    • Suggested starting point: ask questions about your workbook
  2. 0:34 – 1:04

    Budget checks and quick analysis from existing workbook data

    Claude answers questions by scanning the sheet and performing calculations that would otherwise be done manually. The example checks whether travel and meals stay under 40% of total expenses and shows the math for verification.

    • Claude reads the expense report and calculates totals
    • Adds up specific categories (travel + meals) and computes % of spend
    • Compares results against a target threshold (40%)
    • Shows the underlying math so you can validate the result
  3. 1:04 – 2:04

    Debugging spreadsheet errors by tracing formulas to root causes

    Claude helps diagnose Excel errors by interpreting error messages and inspecting the workbook context. In the sales sheet example, it pinpoints a division-by-zero caused by missing units data.

    • Identify where an error occurs (average price in May)
    • Trace formula dependencies to locate the source cell
    • Explain the precise cause (revenue/units with empty units cell)
    • Recommend practical fixes (fill missing data)
  4. 2:04 – 2:34

    Explaining opaque formulas (VLOOKUP breakdown with citations)

    For inherited or complex spreadsheets, Claude can explain formulas step-by-step. The example demystifies a VLOOKUP-based letter-grade conversion and provides citations to jump to referenced cells.

    • Break down a VLOOKUP formula into its components
    • Explain lookup value, table array, column index, and match type
    • Clarify what TRUE means (approximate match)
    • Provide citation boxes to navigate directly to referenced cells
  5. 2:34 – 3:37

    From Q&A to multi-step automation: cleaning messy raw sales data

    The workflow shifts from single questions to multi-step tasks. Claude is prompted to standardize dates, remove duplicates, sort records, and create an annual summary sheet—requesting permission before making changes.

    • Handle multi-step data cleaning with a single clear prompt
    • Standardize date formats to YYYY-MM-DD
    • Remove duplicates and sort chronologically
    • Create a new summary sheet (Historical) with annual totals and counts
    • Ask for permission before modifying the workbook
  6. 3:37 – 4:09

    Adding interpretation and context to cleaned data (trend insights)

    After performing transformations, Claude summarizes what it changed and adds analytic commentary. It highlights year-over-year growth and caveats like partial-year data affecting interpretation.

    • Report actions taken (standardization, duplicates removed, sorting)
    • Summarize annual revenue totals and transaction counts
    • Call out growth (e.g., 2024 up vs 2023)
    • Note limitations such as incomplete year data (only 10 months recorded)
  7. 4:09 – 4:40

    Building a dynamic multi-year forecast with adjustable assumptions

    Claude creates a three-year forecast model driven by an assumptions sheet. Forecast values link via formulas to the growth-rate assumption so users can easily adjust scenarios.

    • Create Forecast and Assumptions sheets to separate inputs from outputs
    • Calculate an average annual growth rate and use it for projections
    • Project future years (2026–2028) using linked formulas
    • Build a fully dynamic model that updates when assumptions change
    • Flag partial-year history that can skew growth-rate calculations
  8. 4:40 – 5:10

    Finance use case: generating a full DCF valuation model in Excel

    Claude demonstrates building a discounted cash flow (DCF) model from historical financials using specified discount rate and terminal growth assumptions. It outputs implied enterprise value and remains dynamic for scenario testing.

    • Ingest five years of historical financials to infer trends
    • Set key assumptions (10% discount rate, 3% terminal growth)
    • Project free cash flows and compute terminal value
    • Discount cash flows back to present to derive enterprise value
    • Adjust assumptions and see the valuation update automatically
  9. 5:10 – 5:40

    Creating pivot tables automatically from transaction-level data

    Claude tackles a common Excel pain point: building pivot tables. Using a sales transactions dataset, it generates a pivot showing total revenue by region and product.

    • Use transaction data (date, region, product, rep, revenue) to build pivots
    • Create a pivot table with regions as rows and products as columns
    • Aggregate to total revenue automatically
    • Replicate a manual pivot workflow with less setup effort
  10. 5:40 – 6:10

    Turning pivots into charts and iterating on chart design via prompts

    Claude creates a chart from the pivot table and then edits it based on follow-up instructions. It can change chart type, labels, titles, axes, and styling while explaining changes made.

    • Generate a chart directly from the created pivot table
    • Select an appropriate chart type and add labels
    • Iterate via prompts (e.g., change to stacked bar, add a title)
    • Edit chart components: axes, title, colors, and type
    • Explain modifications and reasoning for transparency
  11. 6:10 – 6:40

    Transparency, verification, and data-handling cautions

    The video emphasizes that Claude shows its work and provides citations to the underlying cells—tools you should use to verify outputs. It also reminds viewers to follow organizational standards and policies, especially with sensitive or regulated data.

    • Review Claude’s step-by-step changes and reasoning
    • Use citation boxes to jump to referenced cells for validation
    • Verify outputs against internal methodologies before sharing externally
    • Follow organizational policies for sensitive/regulatory data handling
  12. 6:40 – 7:09

    Wrap-up: faster spreadsheet work by shifting focus to decisions

    The conclusion reinforces the value proposition: Claude reduces mechanical spreadsheet labor so users can focus on assumptions, scenarios, and interpretation. It ends by repeating how to open the Claude sidebar in Excel.

    • Claude accelerates work across finance, accounting, operations, and general Excel use
    • Automates mechanical tasks to free time for decision-making
    • Encourages focusing on assumptions, scenarios, and narrative
    • Reminder of shortcut to open Claude (Mac/Windows)

Get more out of YouTube videos.

High quality summaries for YouTube videos. Accurate transcripts to search & find moments. Powered by ChatGPT & Claude AI.