Formatting Data for Upload

Format options

Before you upload your own data into mySidewalk, it’s important to make sure your file is clean, clear, and well-structured. Formatting your data correctly upfront will help the platform interpret it accurately, speed up your upload process, and unlock better insights.

To help you out, we’ve created a two-part series that walks through the most common formats:

Both guides include examples, tips for cleaning up your data, and even prompts you can copy/paste into AI (like ChatGPT or Claude) to help with formatting.

 

Accepted file types

In addition to format, you should also select a filetype to upload. CSV is what we recommend.

mySidewalk supports uploading CSV, GeoJSON, SHP, KML, GML, and GDB. But CSVs are what allow you to take full advantage of mySidewalk's features.

Why use CSV?

Uploading as a CSV enables you to match the places in your file with the places in our library. That means you can:

---

Option 1: Files with Geographic Information

Let’s walk through how to prepare your file so that each row represents a place (like a county), and each column contains a specific variable and time period. You’ll also learn how to ask an AI assistant to help clean things up.

Proper formatting leads to smoother uploads, accurate maps, and clearer insights—getting it right from the start saves time and ensures better results.

 

Step 1: One row = one place

Each row in your file should represent a single location (ZIP, city, county, etc.).

 

Step 2: Time belongs in the column name

Don't use a separate “Year” column. Instead, include the year (or other time period) in the column name.

Remember to use consistent time formatting. Examples:

• Grants_2020
• Grants_Jan_2021
• Grants_2021_q2

 

Step 3: Include breakdowns as separate columns

If you have subcategories (e.g., executive vs. non-executive payroll), use clear, separate column names:

 

Step 4: Clean before upload

✅ Do: 

• Only one header row

🚫 Don’t: 

• Don't put “Year” in its own column
• Don't include extra header rows or totals as separate rows

 

Step 5: Ask AI to help you reformat (optional)

If your file is messy or structured differently, you can ask an AI to reformat it.

Try something like:

• “I have a spreadsheet where each row is a county, and columns are ‘2020 Grants’, ‘2021 Grants’, ‘2021 Payroll’. Can you help me rename the columns to use underscores, like ‘Grants_2020’?”

Or:

• “My file has rows for each combination of county and year, with time in a column. Can you pivot it so each column becomes a variable like ‘Grants_2020’?”

 

Bringing it all together: a final example

Now you’re ready to upload!

 

---

Option 2: Files without Geographic Information

If your data isn’t tied to multiple geographic areas—but still includes multiple variables and time periods—you should use a wide format.

In this structure, each row represents one subject (like an organization or summary report), and each column represents a specific variable + time combination.

To help interpret your data more easily, include a “Variable Label” column as the first column. This column gives a short, human-friendly name or description of the dataset or subject.

 

Step 1: One row = one subject or dataset

Each row represents a single dataset summary or reporting entity. The “Variable Label” column gives a descriptive label for the row.

Example

Step 2: Time belongs in the column name (optional)

If your data file contains a time element, each column header should include both the variable name and time period.

Example

Variable Label	Grants_2019	Payroll_2020
Summary Totals	2800000	3000000

 

Step 3: Add more variables and breakdowns

Each variable and time combo gets its own column. You can expand this structure by adding other variable types (like Payroll), and use consistent naming.

Example

Variable Label	Grants_2019	Grants_2020	Payroll_2019	Payroll_2020
Summary Totals	2800000	3000000	2200000	2400000

Example with breakdowns

Variable Label	Grants_2019	Payroll_Total_2020	Payroll_Exec_2020	Payroll_NonExec_2020
Summary Totals	2800000	2400000	900000	1500000

 

Step 4: Clean before upload

✅ Do:

• Use a Variable Label column to provide context while you upload
• Use one row per subject/dataset

🚫 Don’t: 

• Use separate rows for each variable or time period
• Put the year in a separate column
• Mix naming styles (e.g. Pay)Exec_2020 vs. Payroll_Exec_2020)

 

Step 5: Ask AI to Help You Reformat

If your file is in long format (e.g. one row per variable per year), you can ask the AI assistant to help restructure:

• “Can you pivot this long-format table into a wide format with one row per dataset, one column per variable-year, and include a variable label at the beginning of each row?”

Now your file is fully prepped for upload into mySidewalk!

 

---

Need Additional Help? 

• Download the Upload Template to view examples
• Click “I need help” or chat with us (click Support in the main menu, then Chat with us)