4 Prompts That Turn Any CSV Into an Excel Dashboard (Claude in Excel Tutorial)

  • Beginner
  • 6 steps
  • By The Rundown University

Learn four copy-paste prompts that turn a raw CSV into a clean, visual Excel dashboard using Claude in Excel. You will import a dataset, have Claude inspect and clean it, then generate summary tables, conditional formatting, and charts. The cleaning pass can take a minute or two depending on the data, and the dashboard-building step takes about five minutes.

Step by step 6 steps

  1. Step 1 of 6

    Install the add-in and import your data

    Open Excel, click Add-ins in the top ribbon, then Get Add-ins. Search for Claude and install it. The Claude button will appear in your ribbon. Click it and log in.

    Open your CSV in Excel. If it opens as a flat file, save it as a workbook (.xlsx).

    Pro tip: Always duplicate your data into a second tab or save a backup copy before you start. Claude can modify cells directly and occasionally overwrite things, so it is better to have a safety net.

  2. Step 2 of 6

    Explore the data

    Paste this prompt into Claude:

    Look at all the data in this workbook. Tell me what this data is, what each column represents, and flag any problems with the structure or formatting. List the issues.

    Claude will scan the workbook and return a plain-English breakdown of what the data is, what each column means, and what is wrong with the structure or formatting. This helps Claude understand the dataset before it starts building anything.

    • If Claude gets confused about something, such as abbreviations or date formats, answer its questions in the chat. A one-line clarification goes a long way.
    • You do not need to tell Claude what kind of data this is. It figures that out on its own.

  3. Step 3 of 6

    Clean and structure the workbook

    Paste the second prompt:

    Fix every issue you just identified. Remove any junk rows or metadata. Format numbers so they're human-readable. Create proper tables with clear headers. Sort by whatever the most important metric is.

    Claude will remove junk rows, reformat numbers, build a proper table with headers, and sort by the metric that matters most.

    • This step can take a minute or two depending on how much data you have.
    • You can watch Claude work in real time as it edits cells.
    • If it flags something you disagree with, such as renaming a column, tell it to skip that one and fix the rest.
    Pro tip: Turn on session logging in Claude’s settings. It tracks every change Claude makes in a separate sheet, which is helpful if you need to undo something or explain the process to a teammate.

  4. Step 4 of 6

    Build the dashboard

    Paste the third prompt:

    Create a new "Dashboard" tab. Identify the top 3-5 most useful metrics from this data and build a summary table for each. Rank the entries. Use conditional formatting to highlight the top performers in green and the bottom performers in red.

    Claude creates a new tab and builds summary tables based on the metrics it identifies as most useful. In the source examples, it chose rankings, growth rates, and year-over-year change for GDP data, and scoring leaders, assist leaders, and efficiency for NBA stats.

    • Claude sometimes puts dashboard content in the wrong tab. If that happens, tell it: “I noticed some dashboard tables ended up in the raw data sheet. Can you move everything to the Dashboard tab?”
    • This step takes about five minutes. You can let it run in the background while you do other things.

  5. Step 5 of 6

    Add charts

    Paste the final dashboard prompt:

    Add 2-3 charts to the Dashboard tab that tell the story of this data. Pick the chart types that make the most sense (bar, line, scatter, etc). Make them clean and labeled.

    Claude picks chart types based on the data. For GDP data, it built a line chart showing GDP over time and a bar chart comparing countries. For NBA stats, it made a bar chart for top scorers and a scatter plot for points vs. assists.

    • Charts are a newer feature in Claude for Excel. They work, but Claude sometimes hard-codes source data and then hides the rows, which breaks the chart. If your charts look blank, unhide the rows above or below them.
    • You may need to drag charts around or resize them. The data is right, but the layout may need a nudge.

  6. Step 6 of 6

    Turn the dashboard into a reusable template

    If you want to keep using the dashboard every month or quarter, paste this optional prompt:

    Turn this dashboard into a reusable template. Replace the data with placeholder labels so I can swap in new data next month.

    Claude will turn the dashboard into a template workbook so you can drop in fresh data and run the prompts again.

That's the whole workflow

Mark it complete to keep your progress. The background follows.

Go further with Claude

This Rundown Pro recording covers the same tool or topic as this guide.

$29/mo, cancel anytime

Background and context

Who This Is Useful For

  • Analysts and operations teams who spend the first hour of a project cleaning data
  • Founders and managers who receive CSV exports but are not Excel power users
  • Students and researchers working with messy public datasets

What You Will Build

You will build a polished Excel dashboard from a raw CSV, with reusable prompts you can apply to other datasets.

  • Cleaned and structured workbook data
  • Summary tables based on the most useful metrics
  • Conditional formatting for top and bottom performers
  • Two to three labeled charts on a Dashboard tab
  • A reusable set of four prompts for future CSV files

The workflow starts from messy CSV-style data like this and turns it into a dashboard.

What You Need

  • Microsoft 365 with Excel on desktop, web, or Mac
  • A Claude Pro, Max, Team, or Enterprise plan for the Excel add-in
  • A CSV file, such as your own data, a client export, or a public dataset

Going Further

  • Try the same four prompts with your own data from a CRM, accounting tool, ad platform, or another source.
  • Build template workbooks for recurring team use, such as monthly reports, client dashboards, or inventory tracking.
  • Have Claude convert a finished dashboard into a template so you can reuse it every cycle with fresh data.