Connect Google Sheets to a Public CSV Feed—and Know What Can Break

To import a public CSV URL into Google Sheets, select an empty cell and enter:

=IMPORTDATA("https://example.com/data.csv")

This approach is best suited to a URL that returns CSV or TSV data directly without requiring sign-in. The sheet remains dependent on that external resource, so verify the imported structure and do not assume immediate updates or permanent availability.

Use IMPORTDATA to load the public CSV

Google documents IMPORTDATA as the native Sheets function for importing CSV- or TSV-formatted data from a URL. The URL must include its protocol, such as https://, and a URL written directly in the formula must be enclosed in quotation marks:

=IMPORTDATA("https://example.com/data.csv")

You can also supply the URL through a cell reference, as shown in the official Google Sheets IMPORTDATA documentation.

Because the result expands down and across from that cell, use an otherwise empty output area with enough available rows and columns.

For a more maintainable setup, place the feed URL in a separate cell—for example, Source!A1:

https://example.com/data.csv

Then enter the formula on another sheet or in a separate output area:

=IMPORTDATA(Source!A1)

A cell reference makes the source easier to inspect, replace, and document without editing the import formula.

What matters is that the resource returned at that URL contains compatible CSV- or TSV-formatted data.

For a functioning practice source, one third-party tutorial uses this public CSV URL:

https://raw.githubusercontent.com/mwaskom/seaborn-data/master/iris.csv

Place that URL in Source!A1, enter =IMPORTDATA(Source!A1) in an empty sheet, and inspect the resulting headers and rows. Use a small public or synthetic dataset for practice rather than operational or confidential records.

Check the URL before connecting the sheet

Run this four-item preflight check before adding a feed:

1. The URL includes http:// or https://.
2. It returns CSV or TSV data directly, not an ordinary webpage or preview.
3. It works without sign-in or access tied to your browser session.
4. It contains no confidential records, API keys, access tokens, or other credentials.

Open the URL in a private or incognito browser window. This can reveal obvious login requirements or permissions inherited from your normal browser session. A usable resource should expose or download the underlying delimited data rather than present an application interface.

An HTML preview, file-edit page, or ordinary sharing page is not necessarily the data resource itself. This distinction is particularly important with file-sharing services: a page may display a CSV filename and preview its contents while still returning HTML rather than CSV to a requester.

The private-window test is useful but not conclusive. A community support report documents a CSV URL that continued to work directly while Google Sheets returned “Could not fetch url,” illustrating that browser access alone does not guarantee successful retrieval by IMPORTDATA (Metabase discussion).

Do not make private data public simply to use this formula. Likewise, do not embed credentials in a URL that will be plainly visible in a spreadsheet formula. This workflow is intended for data already approved for anonymous public access.

Work with the imported table

A practical workbook structure separates the source, raw import, and analysis:

1. Create a sheet named Source.
2. Put the public CSV URL in Source!A1.
3. Create a sheet named Raw data.
4. Enter this formula in Raw data!A1:

=IMPORTDATA(Source!A1)

Reserve the Raw data sheet for imported output. Put calculations, labels, charts, and corrections elsewhere so they do not block the expanding table or become mixed with source records.

To create a short preview, wrap the import in QUERY:

=QUERY(IMPORTDATA(Source!A1), "select * limit 10")

Third-party instructional examples demonstrate combining QUERY with IMPORTDATA to limit, select, or filter the result shown in Sheets.

Do not interpret that formula as evidence that Sheets retrieves only those rows from the upstream CSV. The supplied documentation establishes that QUERY can restrict the resulting table, but not that it reduces the source file fetched. If a feed is too large or costly to retrieve, provide a smaller source file before connecting it to Sheets.

Before relying on the imported table, inspect:

• header names and the number of header rows;
• row and column alignment;
• delimiter handling;
• identifiers that may have lost leading zeros;
• dates that may have been reinterpreted;
• blank values and missing-value markers;
• quoted fields containing commas or line breaks;
• number, currency, and percentage formatting.

Keep the imported area unchanged when the feed will be reused. Perform cleaning and transformations in a separate sheet or through a documented process.

A URL-connected formula also differs from a one-time CSV upload. Use the formula when source changes are useful; use a one-time copy when you need a fixed input. A current third-party overview describes these URL-based and uploaded CSV workflows.

Understand refresh and availability limitations

Treat IMPORTDATA as an external dependency, not as a guaranteed real-time delivery system. The cited Google help page defines the function’s syntax and accepted formats, but it does not specify a refresh interval. Third-party sources make inconsistent claims about timing, so do not build a time-sensitive publication around an assumed cadence or promise that a source change will appear immediately.

A recurring sheet also depends on the publisher keeping the resource available and compatible. The connection may stop working if:

• the URL moves or the file is deleted;
• anonymous access changes;
• the publisher’s server is unavailable;
• the resource returns an error;
• retrieval controls change;
• repeated requests are restricted.

For recurring reports, check expected headers, column order, identifier formats, units, and any source-provided update date before publishing new results. If downstream formulas use column positions, document those mappings and review them whenever the source structure changes.

When auditability or reproducibility matters, preserve a dated copy of the CSV used for the analysis. Record the live feed URL as provenance, but base the published work on the saved input.

Troubleshoot a CSV URL that will not import

Start with the source URL, returned format, and available output area rather than repeatedly deleting and re-entering the formula.

Symptom	Checks	Next action
Resource at URL not found	Check for a typo, obsolete endpoint, moved file, or deleted resource.	Locate the publisher’s current direct data URL and update the source cell.
Could not fetch URL	Test anonymous access, availability, and whether the resource returns CSV or TSV rather than HTML.	Retry after a temporary outage or obtain a stable direct resource from the publisher.
Result will not expand	Check whether values, formulas, or insufficient columns obstruct the output area.	Move the formula to a clear sheet or remove the obstruction.
Implausible number of columns	Inspect the delimiter and raw response. Check whether the URL returned a webpage instead of data.	Correct the source or use a publisher-supported CSV resource.
Drive view or edit page supplied	Determine whether the URL opens a sharing interface rather than the CSV itself.	Obtain a documented direct data URL instead of relying on the ordinary sharing page.
Headers or calculations changed	Compare the current headers, order, units, and field definitions with the previous version.	Pause publication until the schema and downstream mappings are confirmed.

Community posts contain various Google Drive download-link patterns, but they do not establish a current, universal interface or guarantee. Prefer a publisher-documented export resource or an approved public file hosted at a stable direct URL.

A successful browser test still does not prove that Sheets can retrieve the resource. Conversely, “Could not fetch URL” does not identify one definitive cause. Check the URL, anonymous availability, returned content, and source status without assuming a particular response or redirect behavior.

Do not promise that re-entering the formula or using an alleged refresh command will bypass caching or force an immediate update. The supplied official documentation does not establish such behavior.

Choose a live feed or a documented snapshot for publication

Choose the source method according to the needs of the finished work:

Choose this approach	When it fits
Live IMPORTDATA connection	Source changes are useful, exact update timing is not critical, and someone can monitor availability and structural changes.
Dated CSV snapshot	Readers need a reproducible record, the work must be auditable, or later source revisions could alter the findings.

Before analyzing or republishing the table, record:

• the source publisher;
• the direct feed URL;
• the retrieval date;
• field definitions and units;
• the publisher’s stated reuse or license terms;
• filters, corrections, joins, and other transformations.

Public access and reuse terms are separate checks. Record the terms supplied by the publisher and resolve unclear permissions before redistributing the dataset. Validate the table, preserve identifiers, normalize missing values, and explain field meanings for readers who did not build the spreadsheet.

Once the spreadsheet is clean, documented, and free of sensitive information, TablePage is one downstream option for turning it into a public interactive data page. Publishing is a separate step: a successful import does not establish that the data is accurate, adequately documented, or ready for release.

Before publication, confirm that:

• the publisher, source URL, and retrieval date are recorded;
• headers and downstream column mappings have been checked;
• field meanings, units, and transformations are documented;
• the table contains no sensitive or confidential information;
• stated reuse terms have been reviewed;
• a dated input file has been preserved when reproducibility matters.